Biomechanics

BIOMECHANICS: THEORY AND APPLICATIONS SERIES
BIOMECHANICS: PRINCIPLES, TRENDS

AND APPLICATIONS
No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or
by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no
expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No
liability is assumed for incidental or consequential damages in connection with or arising out of information
contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in
rendering legal, medical or any other professional services.
BIOMECHANICS: THEORY AND
APPLICATIONS SERIES
Biomechanics: Principles, Trends and Applications

Jerrod H. Levy (Editor)
2010. ISBN: 978-1-60741-394-3
BIOMECHANICS: THEORY AND APPLICATIONS SERIES
BIOMECHANICS: PRINCIPLES, TRENDS

AND APPLICATIONS
JERROD H. LEVY
EDITOR
Nova Science Publishers, Inc.

New York
Copyright © 2010 by Nova Science Publishers, Inc.
All rights reserved. No part of this book may be reproduced, stored in a retrieval system or
transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical
photocopying, recording or otherwise without the written permission of the Publisher.
For permission to use material from this book please contact us:
Telephone 631-231-7269; Fax 631-231-8175
Web Site: http://www.novapublishers.com
NOTICE TO THE READER

The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or
implied warranty of any kind and assumes no responsibility for any errors or omissions. No
liability is assumed for incidental or consequential damages in connection with or arising out of
information contained in this book. The Publisher shall not be liable for any special,
consequential, or exemplary damages resulting, in whole or in part, from the readers‘ use of, or
reliance upon, this material.
Independent verification should be sought for any data, advice or recommendations contained in
this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage
to persons or property arising from any methods, products, instructions, ideas or otherwise
contained in this publication.
This publication is designed to provide accurate and authoritative information with regard to the
subject matter covered herein. It is sold with the clear understanding that the Publisher is not
engaged in rendering legal or any other professional services. If legal or any other expert
assistance is required, the services of a competent person should be sought. FROM A
DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE
AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS.
LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA
Available upon request.
ISBN: 978-1-61761-865-9 (Ebook)
 New York
CONTENTS
Preface vii
Chapter 1 Arts Biomechanics – An Infant Science: Its Challenges and Future 1
Gongbing Shan and Peter Visentin
Chapter 2 Understanding Corneal Biomechanics through Experimental
Assessment and Numerical Simulation 57
Ahmed Elsheikh
Chapter 3 Biomechanics Concepts of Bone-Oral Implant Interface 111
Ahmed Ballo and Niko Moritz
Chapter 4 Biomechanical Remodeling of the Diabetic Gastrointestinal Tract 137
Jingbo Zhao, Donghua Liao, Jian Yang and Hans Gregersen
Chapter 5 Biomechanics of the Gastrointestinal Tract in Health and Disease 163
Jingbo Zhao, Donghua Liao and Hans Gregersen
Chapter 6 Electromyography in the 21st Century: From Voluntary Signals
to Motor Evoked Potentials 207
Petra S. Williams and Brian C. Clark
Chapter 7 Biomechanics in Children with Cerebral Palsy 233
Jessie Chen and Dinah Reilly
Chapter 8 Biomechanical Properties of Cornea 251
Sunil Shah and Mohammad Laiquzzaman
Chapter 9 Some Aspects of the Biomechanics of Skilled Musical Performance 267
Jessie Chen and George Moore
Chapter 10 Contact Hip Stress Measurements in Orthopaedic Clinical Practice 281
Blaž Mavčič, Matej Daniel, Vane Antolič, Aleš Iglič
and Veronika Kralj-Iglič
Chapter 11 External Pelvic Fixation during Lumbar Muscle Resistance Exercise 295
Michael C. McGlaughlin, Philip A. Anloague and Brian C. Clark
vi Contents
Chapter 12 Bone Cell Adhesion: An Important Aspect of Cell Biomechanics in

the Development of Surface Modifications for Orthopaedic
Implants 305
Andreas Fritsche, Frank Luethen, Barbara Nebe,
Joachim Rychly, Ulrich Lembke, Carmen Zietz,
Wolfram Mittelmeier and Rainer Bader
Chapter 13 The Differences in Biomechanical Patterns of Fast Motor Learning
of Children and Adults 315
A. Skurvidas, A. Zuoza, B. Gutnik and D. Nash
Chapter 14 Applying Pressure Sensors and Size Differences in Running Shoes
Fit Measurement 317
Y. L. Cheng and Y. L. Hong
Chapter 15 Improvement of the Input Data in Biomechanics: Kinematic and
Body Segment Inertial Parameters 351
Tony Monnet, Mickael Begon, Claude Vallee
and Patrick Lacouture
Index 385
PREFACE
Biomechanics is the application of mechanical principles (statics, strength of materials

and stress analysis to the solution of biological problems of living organisms. This includes
bioengineering, the research and analysis of the mechanics of living organisms and the
application of engineering principles to and from biological systems. This research and
analysis can be carried forth on multiple levels, from the molecular, wherein biomaterials
such as collagen and elastin are considered, all the way up to the tissue and organ level. This
new and important book gathers the latest research from around the globe in the study of this
dynamic field with a focus on issues such as; art biomechanics, understanding corneal
biomechanics, biomechanical remodeling of the diabetic gastrointestinal tract, biomechanics
in children with cerebral palsy, cell biomechanics for orthopaedic implants, and others.
Chapter 1 - While biomechanics has achieved successes in many fields involving
locomotion, motor learning, skill acquisition, technique optimization, injury prevention,
physical therapy and rehabilitation, one area has heretofore been scarcely represented in the
literature – Arts Biomechanics. Biomechanics clearly has significant potential for application
in the performance arts, such as music and dance, since skills needed for these activities are
visibly related to the human musculoskeletal and nervous systems. In such areas, Arts
Biomechanics should begin by focusing on skill analyses and acquisition necessary for the
performance of the artistic act. Subsequently it should engage in a deeper discourse that
explores the relationship between these and the desired aesthetic outcome. Less apparently,
biomechanics may also enhance the analysis and comprehension of other arts, such as
painting, where gesture is often embedded in the artwork by means of symbolism, tradition,
the process of art creation, or as an inherent product of the existential nature of humanity.
Chapter 2 - The Ocular Biomechanics Group was established in 2002 with one clear
target; to develop a virtual reality model of the human eye that can be used effectively and
reliably to predict ocular response to surgery, injury and disease. This ambitious, and
seemingly illusive, target helped plan our activities over the last 6 years and will still be
focusing our efforts as the authors strive to create the necessary knowledge using
experimental methods, build the predictive tools using programming and analysis means, and
validate the findings in both the laboratory and the clinic. This chapter presents an overview
of our biomechanical studies from laboratory material characterisation to finite element
numerical simulation. The chapter describes what has been achieved and points at the
remaining gaps in our knowledge. It explains that while much remains unknown in ocular
behaviour, the authors are now in a good position to use available knowledge to progress
viii Jerrod H. Levy
predictive modelling and use it in actual applications such as improving the accuracy of
tonometry techniques, planning of refractive surgeries and design of contact lenses. The
discussion focuses on the cornea, although scleral biomechanics receive some mention. The
chapter also refers to microstructural, biomechanical and topographic studies conducted by
other research groups. Coverage of these studies has been necessary to provide a more
complete image of current understanding of corneal biomechanics.
Chapter 3 - Osseointegrated implants are actually replacements for natural teeth, and, like
natural teeth, they are exposed to various forces. The success of osseointegration is based on
the clinical outcome; clinicians must ensure that the stresses that the superstructure, implant,
and surrounding bone are subjected to are within the tolerable limits of the various
components. Structural compatibility is the optimum adaptation to the mechanical behavior of
the host tissues. Therefore, structure compatibility refers to the mechanical properties of the
implant material, such as elastic modulus, strength, implant design and optimal load
transmission (minimum interfacial strain mismatch) at the implant/tissue interface, which is
the key to the successful functioning of the implant device.
This chapter reviews some of the reaction, properties and characteristics of the bone and
explains how the bone-implant interface will react under loading condition. The chapter also
includes characteristics, properties and other important information about the implant
biomaterials and implant coating.
Chapter 4 - Gastrointestinal tract sensory-motor abnormalities are common in patients
with diabetes mellitus with symptoms arising from the whole GI tract. Common complaints
include dysphasia, early satiety, reflux, constipation, abdominal pain, nausea, vomiting, and
diarrhea. The pathogenesis of GI symptoms in diabetes mellitus is complex in nature, multi-
factorial (motor dysfunction, autonomic neuropathy, glycemic control, psychological factors,
etc.) and is not well understood. Histologically, many studies have demonstrated prominent
proliferation of different GI wall layers during diabetes. During the past several years, several
studies demonstrated that experimental diabetes induces GI morphological and biomechanical
remodeling. Following the development of diabetes, the GI wall becomes thicker and the
stiffness of the GI wall increases in a time-dependent manner. It is well known that
mechanosensitive nerve endings exist in the GI tract where they serve a critical role for tissue
homeostasis and symptom generation. Mechanoreceptor-like structures such as
intraganglionic laminar nerve endings and intramuscular arrays have been identified. The
changes of stress and strain in the GI wall will alter the biomechanical environment of the
mechanosensitive nerve endings, therefore, the structure as well as the tension, stress and
strain distribution in the GI wall is important for the sensory and motor function.
Biomechanical remodeling of diabetic GI tract including alterations of residual strain and
increase in wall stiffness will alter the tension and stress distribution in the vicinity of the
mechanosensitive afferents with consequences for perception and motility of the GI tract.
Chapter 5 - The gastrointestinal (GI) tract is functionally subjected to dimensional
changes. Hence, biomechanical properties such as the stress-strain relationships are of
particularly importance. These properties vary along the normal GI tract and remodel in
response to growth, aging and disease. The biomechanical properties are crucial for GI motor
function because peristaltic motion that propels the food through the GI tract is a result of
interaction of the passive and active tissue forces and the hydrodynamic forces in the food
bolus and remodeling of the mechanical properties reflects the changes in the tissue structure
that determine a specific motor dysfunction. Therefore, biomechanical data on the GI wall are
Preface ix
important to understand the pathogenesis to the GI motor-sensory function and dysfunction.

Moreover, biomechanical studies of the GI tract pave the way for further mathematical and
computational modelling. Biomechanical studies of the GI tract will advance our
understanding of GI physiological function, diseases such as dyspepsia and visceral pain, and
GI dysfunction due to systemic diseases. Furthermore, integrated GI simulation models will
be beneficial for medical education and for evaluation of the efficacy and safety of new drugs
on GI function.
Chapter 6 - The force produced by skeletal muscle is controlled by the electrical signals
being sent from motor neurons to muscle fibers. These electrical signals, which are known as
action potentials, can be recorded as they travel along the muscle cell membrane and are
referred to as an electromyogram (EMG) signal. It has been more than a century since the
first recording of a voluntary EMG signal was reported, and today it has become a classic
technique for evaluating and recording the activation of skeletal muscles during human
movement. In recent years, the advent and development of transcranial magnetic stimulation
has re-invigorated EMG research, and it is now possible to safely and painlessly evoke EMG
signals directly from the motor cortex of conscious humans. This chapter reviews the
recording and measurement issues associated with EMG and its respective applications.
Particular attention is paid to its role in understanding the neuromechanics of human
movement.
Chapter 7 - Children with cerebral palsy (CP) lack the higher-level motor skills present in
age-matched typically developing (TD) children. The development of postural control is
critical to the acquisition of increasingly complex motor skills as well as to the production of
coordinated motor behavior, such as locomotion. This chapter examines recent developments
in understanding the abnormal postural control in children with CP and assessments of the
effectiveness of rehabilitation techniques using biomechanics measurements. The authors
show that the delayed and impaired development of postural control in children with CP is
not only due to the immaturity of central nervous system but also abnormal postural
alignment and muscle force production.
Children with CP often have difficulty maintaining stability when facing unexpected
threat to balance. The authors present studies comparing reactive balance control in children
with spastic diplegic cerebral palsy (SDCP) and TD children using support surface
perturbation and show that a number of factors contribute to poor balance control in children
with SDCP. 1) There was a temporal disorganization of joint torque activation. 2) There was
a spatial disorganization of the joint torque profiles. 3) Children with SDCP also showed
slower speed to reach peak torque value. In addition, the authors show that when TD children
were asked to mimic crouched stance as that seen in children with SDCP, they exhibited
abnormal postural control as well, indicating that musculoskeletal constraints are also
contributors to the atypical postural muscle response patterns seen in children with SDCP.
These findings suggest that the neuromuscular response patterns of some children with SDCP
may be appropriate strategies for their musculoskeletal constraints secondary to deficits in the
neural system.
In this chapter the authors also discuss the culmination of our findings in relation to
clinical applications in the management of musculoskeletal impairments to improve postural
control in children with SDCP, and the significance of using biomechanical measures to show
a direct relationship between the impairments of the musculoskeletal system and reactive
postural control as well as possible coping strategies used by children with SDCP. The
x Jerrod H. Levy
authors examine the current biomechanical research used to ascertain the effectiveness of two
therapeutic interventions purported to affect the musculoskeletal system for the improvement
of function in children with SDCP, ankle foot orthoses (AFO) and strength training.
Finally, the authors examine the gaps in current clinical research when assessing the
effectiveness of interventions to reduce the musculoskeletal impairments constraining static,
reactive, and dynamic balance control in children with SDCP.
Chapter 8 - The knowledge of corneal biomechanical properties of cornea has gained
importance in recent years. Investigators have been trying to find easy and practical ways to
establish these biomechanical properties but to date have had to rely on corneal thickness
measures to give an idea of corneal biomechanics. This review explores what is known about
the biomechanical properties of the human cornea.
An overview of corneal thickness measurements, its impact on measurement of
intraocular pressure and its importance in various disease states is discussed. The recent
advent of the Ocular Response Analyser, an in-vivo measure of ocular hysteresis and corneal
resistance factor and the pulse waveform associated with this will be discussed. The
importance of this machine with respect to corneal biomechanics will be presented.
Chapter 9 - This chapter addresses a fundamental aspect of musical performance in string
players: how the physical ―geography‖ of the instrument and bow, and the anthropometric
dimensions of the player interact to produce the stereotypic motor behavior.
Several factors determine movement, but, unlike most biomechanical tasks, the
determining outcome here is acoustic. Both upper extremities are involved in the performance
but in very different ways.
The left arm governs the contact position of the fingers on the string, and hence the pitch
of a note. The spatial relations of the instrument to the body, the contact point of the finger on
the string, the length of the fingerboard, and the dimensions of the arm determine a unique
posture, and thus the muscle activation patterns, for the left arm for each individual player.
The left arm has quite different postures in relation to the body and the instrument for cello
players in comparison to violinists.
The bowing (right) arm draws the bow across the string. Its travel velocity is the principal
determinant of loudness, but its distance from the bridge, the contact position along the bow
length, bow pressure, and its angle of attack on the string leave a noticeable effect on the tone,
or timbre. Bowing movements are essentially determined by the flexion and extension of the
elbow, with subtle motions of the shoulder and wrist to keep the bow moving in a straight line
perpendicular to the string. Given a fixed spatial relation between the body of the performer
and the instrument, the posture of the arm for a given bow/string contact point is uniquely
determined.
In addition, the force of gravity plays a role in the control of movement. The authors
show, however, that it affects cellists and violinists in very different ways. Left arm
movements of cellists are more affected by gravity than those of violinists; whereas gravity
affects the right arm more in violinists.
In this chapter the authors focus on a series of specialized topics: 1) control of the left
arm during shifting movements; 2) fine control of the left arm during corrections of intonation
errors; 3) coordination between the upper arm and forearm; and 4) coordination between the
left and right arms.
Chapter 10 - There exist several invasive and noninvasive methods to measure the
contact hip stress but due to their complexity only few have so far been tested in clinical trials
Preface xi
with large numbers of participating subjects. Consequently, the use of contact hip stress
measurements in orthopaedic clinical practice is still in its experimental phase.
Biomechanical studies of human hips based on the analysis of 2-D pelvic radiographs
have turned out to be a reasonable compromise between the measurement accuracy and the
feasibility in clinical setting. Clinical studies have shown significantly higher values of hip
stress in adult dysplastic hips when compared to normal hips. It has been found that the
cumulative hip stress independently predicts the WOMAC score after 29-years of follow up
in dysplastic hips and does so better than morphological radiographic parameters of hip
dysplasia or the resultant hip force alone. The preoperative value of the contact hip stress and
the magnitude of its operative correction have been found predictors of the long term success
of the Bernese periacetabular osteotomy. Elevated shear stress in femoral neck, but not
elevated hip contact stress, has been found to be a risk factor for slipping of the capital
femoral epiphysis. A statistically significant correlation between the contact hip stress and the
age at the total arthroplasty has been shown in a group of hips with idiopathic hip
osteoarthritis.
Through advances in 3-D imaging with MRI and CAT scan, visualisation of the femoral
head coverage and pelvic muscle attachment points has improved considerably. However, the
need to supplement the morphological hip status with biomechanical analysis remains. The
current trend is to combine the kinetic gait measurements of the resultant hip force with 3-D
imaging of the hip weight-bearing surface in order to better estimate the contact hip stress for
a given activity/body position. The added value of such measurements over 2-D pelvic
radiograph analysis has not been established yet in clinical trials.
Chapter 11 - Resistance exercise has long been used to promote musculoskeletal health
with the application of training regimens for the clinical treatment and prevention of low back
pain growing in popularity over the last couple of decades. A variety of exercise modes have
been utilized in an attempt to stimulate and promote increases in muscle function of the
lumbar extensors. This chapter examines the current state of knowledge regarding the
application of external pelvic fixation during trunk extension exercise and its importance on
the concomitant increase of functional outcomes such as muscle strength, muscle activation
patterns and compensatory muscle growth.
Chapter 12 - Most revisions of total joint replacements are due to implant loosening,
which is mainly caused by wear particles (―wear disease‖) and inadequate primary implant
stability. The optimised integration of cementless total hip and knee endoprostheses into the
bone stock is the most adequate approach to achieve secondary implant stability and to
prevent implant loosening. Secondary stability is characterized by bone ingrowth of the
implant and decreases the amount of relative implant motion between the implant and bone
stock. It has also been suggested that prostheses which are fully occupied by bone cells are
less susceptible to infection. The economic impact of implant loosening is immense, hence
orthopaedic implant manufactures refine their products continuously.
Many technical developments have improved the survival rate of endoprosthetic
implants. Modern materials and surface modifications such as coatings help to reduce wear
rates, promote cell ongrowth or prevent infections. The cell adhesion of bone cells onto
implant surfaces has not been thoroughly investigated so far. However, different methods to
measure cell adhesion have been described. Some workgroups investigate short-term
adhesion or proliferation of bone cells on implant materials in-vitro, but little is known about
the long-term adhesion. Proliferation or short-term adhesion cannot predict how strong the
xii Jerrod H. Levy
bonding between bone and implant will be. In most cases, cost intensive animal studies have
to be performed in order to gain expressive data. Hence, it is important to assess the bone cell
adhesion forces in an adequate experimental setup in- vitro.
The exploration of bone cell adhesion on surfaces of orthopaedic implants encourages the
development of bio-compatible, bio-active and anti-infectious surfaces. The authors have
developed a test device, based on the spinning disc principle, which allows quantitative
measurements of osteoblastic cells on implant surfaces. First results show differences in
adhesion forces depending on the substrate. In future assessments different bio-active and
anti-infectious surface modifications will be analyzed regarding bone cell adhesion prior to
animal studies.
Chapter 13 – The aim of this chapter was to establish and compare the patterns expressed
on the fast model of motor learning of children and adults executing a fast and accurate task.
The acquisition of a new motor skill follows two distinct stages with continued practice:
first, there is an early, fast learning stage in which performance improves rapidly within a
single training session; later, there is a slower learning stage within the time period of several
sessions of practice. Motor learning is characterized by a specific set of changes in
performance parameters. These changes occur gradually in the course of a learning period.
While the decreases or increases in these parameters have been documented in a variety of
tasks, it remains to be determined whether the time of fast learning is different for children
and adults. Therefore the main aim of this study was to establish if there are differences in
reaction time, average and maximal velocity, trajectory, and accuracy as well as the
variability of these parameters during motor learning. The tasks involved 5 series with 20
repetitions in each.
Chapter 14 – Fit is one of the most critical factors affecting footwear comfort. Blistering,
chafing, bunions and pain may be the result of poor fitting shoes. In long run, it may cause the
foot skeleton deformity.
In order to find out the proper fitting of footwear, it involves getting to know the size of
feet, shoes, and the subjective perception for the shoes selection. Traditional method in
measuring the feet size is to measure the length and width of the feet which can be obtained
easily by tape measure and devices like Brannock. However these are considered to be
insufficient for good footwear fitting. Furthermore, researchers were also encountering
problem in quantifying fit as it is rather subjective which was also suggested to be affected by
shoe wearing experience such as tightness and looseness of the shoes. Therefore researchers
are exploring new method in measuring footwear fit, both objectively and subjectively.
Chapter 15 – Usually, biomechanical models used for human motion analysis are
oversimplified, especially for clinical analyses (Helen Hayes model). The calculated net
joint forces and torques are sensitive to the input data: segment kinematics and body segment
inertial parameters. It is therefore necessary to improve these input data using new methods
and models adapted to the population and movement of interest. The general problem is
divided into three parts: (i) minimization of soft tissue artefacts, (ii) joint centre location and
(iii) identification of the personalized body segment parameters.
In: Biomechanics: Principles, Trends and Applications ISBN: 978-1-60741-394-3
Editor: Jerrod H. Levy, pp. 1-55 © 2010 Nova Science Publishers, Inc.
Chapter 1
ARTS BIOMECHANICS – AN INFANT SCIENCE:

ITS CHALLENGES AND FUTURE
Gongbing Shan and Peter Visentin

1
Department of Kinesiology,
2
Department of Music,
University of Lethbridge,
4401 University Drive, Lethbridge, Alberta. Canada, T1K 3M4
1. OVERVIEW
While biomechanics has achieved successes in many fields involving locomotion, motor
learning, skill acquisition, technique optimization, injury prevention, physical therapy and
rehabilitation, one area has heretofore been scarcely represented in the literature – Arts
Biomechanics. Biomechanics clearly has significant potential for application in the
performance arts, such as music and dance, since skills needed for these activities are visibly
related to the human musculoskeletal and nervous systems. In such areas, Arts Biomechanics
should begin by focusing on skill analyses and acquisition necessary for the performance of
the artistic act. Subsequently it should engage in a deeper discourse that explores the
relationship between these and the desired aesthetic outcome. Less apparently, biomechanics
may also enhance the analysis and comprehension of other arts, such as painting, where
gesture is often embedded in the artwork by means of symbolism, tradition, the process of art
creation, or as an inherent product of the existential nature of humanity.
There are many challenges facing the integration of the Sciences with the Arts. On a
fundamental level, the principles and goals of one often seem at odds with the other. In
reality, neither science nor art is antithetical to the other.
"The most beautiful experience we can have is the mysterious. It is the fundamental
emotion that stands at the cradle of true art and true science. Whoever does not know it and
can no longer wonder, no longer marvel, is as good as dead, and his eyes are dimmed. It was

Ph: (403) 329-2683, e-mail: g.shan@uleth.ca
2 Gongbing Shan and Peter Visentin
the experience of mystery -- even if mixed with fear -- that engendered religion. A knowledge
of the existence of something we cannot penetrate, our perceptions of the profoundest reason
and the most radiant beauty, which only in their most primitive forms are accessible to our
minds…I am satisfied with the mystery of life's eternity and with a knowledge, a sense, of the
marvelous structure of existence -- as well as the humble attempt to understand even a tiny
portion of the Reason that manifests itself in nature."(Einstein, 1931)
However, the self-perceptions of artists and scientists may be problematic. Scientists take
reductionist and reasoned approaches to the world: there is a position/argument; it is
structured in logical steps; a topic sentence names and proves each idea in the discourse; and,
the discourse is aimed at a specific result. For the scientist, these give feelings of comfort,
power, and/or control over the phenomenon. Conversely, the artist is less interested in the
factual and desires to convey the emotional and sensual; reductionist, topical arguments are
considered antithetical to the creation of good art; too precisely formulated a conceptual
structure is perceived as negating mystery and limiting artistic possibilities; and, the result is a
product of the moment – an ever shifting target in time-based performance art. Repeatability
is only a desirable quality in that it can be a measure of skill level. Hence it shows little in
terms of creative ability. The artist wants power and control over others‘ perceptions of the
phenomenon; each individual audient‘s experience of the art responding to and thus
validating the artwork and, by extension, the artist. Perhaps it is the way in which scientists
and artists perceive their own roles that a binary view pervades their respective disciplines –
something is or isn‘t. Given this seeming dichotomy, the question arises ―Why bother with
Arts Biomechanics?‖
The issue that provides the main inertia to expand this nascent field of biomechanics
comes from artists. It is a medical one. Epidemic rates of debilitating injury (48-76%) occur
in performing arts such as music and dance (Brown, 1997; Fry, 1986, 1987, 1988; Fry, Ross,
and Rutherford, 1988; Hagglund, 1996; Hartsell and Tata, 1991; Lockwood, 1988;
Middlestadt and Fischbein, 1989; Zaza, 1992, 1998). Unfortunately, there is currently little
quantitative research examining the aetiology of performance injuries. Existing strategies to
address injuries are largely qualitative and ―experience‖ based. And they are normally
employed only after injury has occurred. Even as artists are becoming increasingly aware
regarding their physical needs, career longevity and injury downtime are becoming pervading
arts industrial issues. Only recently have researchers begun to explore scientific approaches,
such as human performance engineering, to trace causal factors related to human bone,
muscle, and nervous systems injuries in the performing arts (Chesky, Kondraske, Henoch,
Hipple, and Rubin, 2002; W. J. Dawson, 2003, 2007; Solomon and Solomon, 2004). Finally,
artists themselves are slowly turning toward science to provide preventative answers and not
merely remedial ones. From a phenomenological point of view, the performing arts share
many characteristics, including health risks, in common with other skill-oriented activities
(Chesky et al., 2002; Lehmann and Davidson, 2002; Wilson, 1986). Commonalities between
athletic and artistic performance seem obvious, particularly in the area of motor skill analysis,
acquisition during skill learning and performance. Like athletes, elite musicians practice, with
many hours of repetition, to perfect complex motor control sequences. However, music
students are seldom introduced to basic principles of movement science and physiology that
underpin that activity. This is largely because the focus of music teaching is artistic and
outcome driven rather than process oriented – this is a difference between practicing and
Arts Biomechanics – An Infant Science 3
training. While some teachers are knowledgeable regarding efficient use of the body for the
benefit of good performance, many are not. In many parts of the world, classical musicians
today are taught in a manner virtually indistinguishable from that used 50 years ago. Nowhere
in the world could the same be said for elite sport.
As such, it seems logical to build on the successes of Sports Biomechanics in the service
of the Arts. Motor learning, skill acquisition and learning while minimizing injuries constitute
the main research focuses of Sports Biomechanics (Ballreich and Baumann, 1996). However,
there is a fundamental ethical difference between these two fields. Whereas Sports
Biomechanics typically directs is energies to achieving specific, goal-driven, quantifiable
results that are valued for their repeatability (e.g. faster, higher, stronger, etc.), Arts
Biomechanics must be satisfied with guiding the process without appearing to identify an
absolute goal – for reasons of artistic and creative freedom. The emphasis of each is on
training, but Arts Biomechanics needs to contribute to a demystification of the learning and
skill acquisition processes, so that artists can realize their full potential and survive their
chosen vocation.
―My early love affair with dance gradually had been replaced by a struggle to become
what I could not be. I had selected teachers who felt it was their responsibility to tell me over
and over what was wrong, rather than helping me generate the knowledge that would give me
tools to make things right‖ (Evans, 2003)
Clearly, the need to consider artistic values as well as scientific ones creates particular
challenges for the field of Arts Biomechanics. Facing these challenges is, in our opinion, best
served by a multidisciplinary approach – one where research does not simply adopt a
scientific or artistic practice, but engages in discussion that ultimately transcends the
viewpoint of each discipline. Meaningful and relevant research results will be those where
science informs artistry rather than attempting to modify it. Currently, music and dance are
the two dominant areas of biomechanics research in the arts. For this reason, discourse below
will primarily focus on these disciplines. The main effort of this chapter is to summarize the
state of Arts Biomechanics in the following areas: 1) skill analysis, acquisition and pedagogy,
2) injury risk identification, quantification, prevention and compensation strategies, and 3)
Innovative uses of the tools of movement science in the analysis and creation of art. Further,
this chapter will provide discussion that identifies some of the challenges facing Art
Biomechanics, elaborate on its potential and identify some future directions.
2. PERFORMANCE SKILL ANALYSIS AND ACQUISITION

2.1. Historical Overview
Documentation pertaining to instrumental performance and dance has a long history. In

terms of modern classical performance traditions, documentation that incorporates
information on mechanics and motor skills begins as early as the 16th century for instrumental
performance and the 17th century for dance (Hilton, 1997; Kolneder, 1993). Its very existence
can be considered evidence of a general desire to improve overall quality of music/dance
performance, both stylistically and technically. Most of the documents are pedagogical in
nature and they are clearly not targeted at the mature or virtuoso performer. For example, in
his seminal treatise on learning to play the keyboard, Carl Phillip Emanuel Bach clearly states
his intent.
―…keyboard instruction could be improved in certain respects to the end that the truly
good which is lacking in so much music, but particularly keyboard music, might thereby
become more widespread. The most accomplished performers, those whose playing might
prove instructive, are not to be found in such numbers as might perhaps be imagined.‖ (Bach,
1759)
In terms of classical dance, the beginnings of ballet and classical notation are associated
with the Court of Louis XIV of France (1639-1714) (Hilton, 1997). Iconography, dance
manuals and descriptions of the desired aesthetic provide some of the first rudimentary
(bio)mechanics information for analysis and acquisition of dance skills (Figure 1 and 2).
Figure 1. Choreography of a Minuet providing rudimentary instruction on placing the feet. The original
source is from an English dance manual by Kellam Tomlinson, and documents the influence of French
culture in the 18th century.
Figure 2. A choreography of steps from ―Balet de Neuf Danseurs‖ by Feuillet. Paris, 1700.
Collectively, such sources contain fundamental instruction on music/dance and their

contemporary aesthetic goals. They include postural descriptions, instruction pertaining to
skill acquisition such as: fingerings in basic positions, embouchure (for wind instruments) and
bowing techniques (for strings), and they describe the phenomenology of physically
interacting with a musical instrument (how to hold it, the consequences of certain muscular
tensions, tablatures revealing mechanical insights, etc.). Choreographed movement in dance
also create a gestural map for both the performer and dancer, since the music must maintain
aesthetic of bodiliness required in dancing (Bach, 1759; Hilton, 1997; Mozart, 1756; Quantz,
1752; Tromlitz, 1791). Clearly, these point to an awareness of the mechanics of the human-
tool interface, whether that interaction is with an instrument or a physical space.
Further, almost on a century-by-century basis, one can observe a progression to more and
more systematic approaches toward artistic performance. Documentation pertaining to violin
performance can be considered exemplary in this regard. Earliest documentation, from the
16th century, was descriptive, providing relatively little instruction on technique. The 17th
Century included some of the earliest tutors. They contained fundamental postural
descriptions and instruction on music basics: string tunings, fingerings in basic positions,
elementary bowing techniques. Included are some instructions on bow direction, accents, and
attempts to describe articulations and expression. Tablatures (musical fingering charts)
revealed insights into pedagogy. Information on style and technique may also be gathered
from music that incorporated performance indications by the composer. The 18th Century rise
of a middle class and a corresponding increased interest in education lead to maturation of
instruction methods and pedagogy. More methods and music were published. Contents
became more detailed with broader variety of musical examples and more instructions how to
coordinate techniques in order to play expressively. Some compositions dealt with specific
technical difficulties or material for specific skill acquisition.
The 19th Century showed evidence of a paradigm shift (Baillot, 1835). There was a move
from general education and a ―collective consciousness‖ to a focus on individual training and
virtuosity. The period marked the beginnings of music pedagogy as a science.
Figure 3. Position of the right arm and wrist held closer to the body when the violinist plays sitting
down. The neck of the violin is then held slightly lower. Position of the hand, wrist and fourth finger in
extensions (top-right). Forced and improper position of the hand, wrist and fourth finger in extensions
(bottom-right).
Documentation used a new tone emphasizing the development of technique as a means to

achieve the highest degrees of artistry. The focus was on technical training (Figure 3). During
the period, a gradual increase in notion of finding the ―best‖ way to practice can be observed.
Some writers advocated a ―mechanical approach‖ to the instrument with attempts to apply
scientific methods to describe these mechanics. Others began to approach the subject of
performance psychology. Regarding anatomy and physiology, methods to this point may be
summarized as rudimentary. They were experience/practitioner-based with superficial
descriptions of postures or (bio)mechanics (Baillot, 1835). Attempts to include an
understanding of anatomy and physiology were a development of late 19th and early 20th
centuries.
Over the course of the 20th century, there was a continued development of systematic
teaching methods. Also, there was an increasing trend toward learning from the standpoint of
different disciplines: physiology, psychology, (bio)mechanics. Some training exercises were
developed (at times by individuals who may have had little to no ―practical‖ experience
playing music – e.g. medical doctors, physiology professors) that pushed mechanics to
conceptual and physiological limits. As can be seen in Figures 4-6, the body was typically
treated as a machine which could be trained, or programmed, through repeated and structured
movement variations (Hodgson, 1958).
Most 20th-century writings that attempted to document teaching methods of ―successful‖
practitioner/pedagogues failed to deal with fundamental psyco-physiological learning,
biomechanics, or neural control except in superficial terms (G. B. Shan, Visentin,
Wooldridge, Wang, and Connolly, 2007). For example, after mentioning postural or internal
motor control factors in general terms, most typically continue with experience-based
description of techniques and their application in selected passages of repertoire. Thus, they
are written in such a way that only those who have experience with the phenomenon and the
aural tradition being described can sympathize with the descriptions.
Figure 4. Illustration considering the curve which results from the identified ordering of notes, while
using a down bow. The two dotted lines give variations in the track followed by the hand.
Figure 5. The figure shows a possible elliptical path (marked 1, 2, 3, 4), traversed by the heel of the
bow during the playing the identified two notes.
Figure 6. Small loops occurring in the bow-hand that result from string crossings for the identified
scales. String crossings occur on an ―up-bow‖ for the figure on the left and on a ―down-bow‖ for the
one on the right.
Clearly, in the 21st century, there can be a role for the field of biomechanics in terms of
the pedagogical process. Demystification of the mechanics of fundamental skills, and
direction regarding the best way to learn them, is in the best interests of these disciplines –
especially if it can lower injury rates and/or improve the physical health and well-being of
performing artists. Currently, Arts Biomechanics in its infancy as a field of study.
2.2. Challenges Faced by Arts Biomechanics
Notwithstanding efforts to address important artistic concepts through written

documentation, the performance arts has and continues to perpetuate its traditions mainly
through an apprenticeship learning model. Advanced teaching/learning typically occurs in a
one-on-one environment and arts education relies heavily on aural traditions, with
pedagogical frameworks by and large limited by conventions and aesthetic values passed
from teacher to pupil. Participants in such a model typically take great pride in the
provenance of their traditions – often validating their approach by tracing their artistic lineage
from teacher to teacher to one or another of the legendary ―greats‖ of the past. Such a mind-
set frequently inspires a mystical and, in some cases, a near-religious reverence of the teacher,
artistic traditions, concepts, and acquired skills. Nearly everyone who is involved in the
performing arts can, through personal experience, cite anecdotal evidence confirming this
assertion.
In such a system, and since the performative arts are heavily influenced by the need for
an audience, performance and pedagogy are driven by external validation. Frequently, a
teacher‘s quality is judged by the successes of his/her ―prize‖ students. As artistic success or
failure is often determined by micro-increments of ―talent‖, this can lead to an emphasis on
immediacy of skill acquisition, sometimes at the expense of sound learning practices
(Visentin, Shan, and Wasiak, 2008). According to Ruggiero Ricci, one of the 20th century‘s
premier violinists:
―Teachers are ruining a lot of kids. Parents love to say ‗my kid plays the Mendelssohn
Concerto‘ so they take their child away from a teacher who knows better and gives him to one
who will push him. … they are the real criminals in this business‖ (Schwartz, 1983).
Notwithstanding this somewhat arcane approach, numerous examples of success within the
model reinforce the perceived legitimacy of its practices.
However, there are significant problems with such simplistic attitudes and rationales.
First, in order to be scientifically defensible, the ―talent‖ must either be an innate
characteristic of individuals, somehow defined by their genetic makeup, OR it must be a
byproduct learned behaviors and environmental circumstance. At any given point of time, it is
relatively easy to observe levels of skill development and ease of skill acquisition (talent) for
individuals within a group of learners. In one scientific study, subjects were sorted into two
groups based on expert assessments of their ―talent‖ (Levitin, 2006). After following the
groups for several years, results proved discrimination on the basis of ―talent‖ to be a false
determinant of success over the long term. Learning/practice was the ultimate cause of
achievement, and not innate ability. Hence, a focus on the learning process needs to dominate
arts pedagogy.
Further, success as defined by the performance is a measure which oft-times fails to

concern itself with the long-term health and development of the artist. Since the medical
literature shows overwhelming evidence of vocational disaster over the ―long-haul‖ due to
technique-related injuries, there has been an increased interest in improving pedagogical
practice in the performing arts. Past studies have shown that 48% - 76% of professional
musicians need to stop performing for extended periods due to occupational injury (Brown,
1997; Fry, 1986, 1987, 1988; Fry et al., 1988; Hagglund, 1996; Hartsell and Tata, 1991;
Lockwood, 1988; Middlestadt and Fischbein, 1989; Zaza, 1992, 1998). As a result, old
platitudes such as ―the way to Carnegie Hall is practice, practice, practice‖ have evolved into
―the way to Carnegie Hall is practice, rest, practice‖ (Kaplinski, 2003). Further, traditional
practices, such expecting students to engage in long hours of repetitive work often with little
consideration for its efficiency or biological implications, are now recognized as part of the
problem, even if there is a reluctance to changing the mind-set. The performing arts appear to
be ready for more scientific collaborations.
Arts Biomechanics can play an important role in challenging the inertia of tradition that
exists in the performing arts. By supplying quantitative rather than empirical evidence for
improving pedagogy, this sub-discipline of applied biomechanics can improve understanding
of physiological phenomena, motor control, the motor learning process and internal loading
implications for the body during practice and performance. And, with care, it can do so
without being perceived as a threat to the artistic end-product. Ultimately, this is the most
significant practical challenge for Arts Biomechanics.
Unfortunately, there is currently little quantitative research in Arts Biomechanics. One
reason for this may be that both the performing arts and biomechanics are highly specialized
studies each with their own cultures. These cultures may have language barriers (the same
words can have different disciplinary meanings, and hence loading), goal differences
(personal and subjective experience is desired for one and quantitative observation of a
process that leads to specific goals for the other), and/or conflicts between their needs (the
arts require flexible outcomes that accommodate changing tastes/ideas and the sciences desire
outcomes that are essentially quantifyable/immutable). Further, from the standpoint of arts
pedagogy, because of anthropometrical variation between teacher and student, it can be the
tendency for an arts teacher to focus on stylistic outcomes, which in turn obliges the student
to find their own motor control means to that particular end. Arts Biomechanics can inform
the teacher/pupil interface, supplying common ground so that a teacher can adjust his/her
understanding to a student‘s individual anthropometry, and so that the student can better
comprehend the teacher‘s instruction. Biomechanics has great potential to provide linkages
between science and the arts. Clearly, collaborative work is needed in order to bridge
disciplinary divergences. In short, Arts Biomechanics needs to be an integrative force, a
nexus for communication that is not merely coordinative of differing views but accounts for
both perceptual experience and scientific quantification at all stages of the work. This is the
most significant conceptual challenge in the application of biomechanics to the arts.
Finally, the overview of this chapter began by identifying some philosophical challenges
for Arts Biomechanics. Individuals within the arts and the sciences have a vested interest in
maintaining the fundamental precepts and traditions which define the relationship between
discipline and practitioner. For the artist, quantification and transparency represent threats to
individuality and artistry. For the scientist, subjectivity is something that merely has not yet
been objectified. Changing attitudes to foster recognition of the potentially symbiotic
relationship and finding convergences that can link these ethics are the most significant
philosophical challenges for Arts Biomechanics.
In any motor skill endeavor, skill acquisition can be categorized into three primary
stages: cognitive, associative and automotive (Magill, 2001). Clearly, given the cultural
characteristic of the performing arts and applied biomechanics disciplines, the greatest
potential for collaboration between the two seems to lie in addressing the cognitive and
associative stages of the learning process. Such an approach does not challenge artistic
values, and it builds on the strengths of biomechanics as a means to understand and further
motor skill development. Additionally, it has potential to effectively address injury prevention
through quantitative research examining the aetiology of performance injuries. Since most
existing strategies to address injuries are largely ―experience‖ based and they are normally
employed only after injury has occurred, it appears logical to begin with an examination of
pedagogy‘s contribution to this problem. Arts Biomechanics has an important role in this
regard (Visentin et al., 2008). To this end, Arts Biomechanics as a field of study has begun
with the following two questions in mind: ―can we identify, describe, and quantify qualities
that lead to artistic success while minimizing the potential of performance injury?‖ and ―how
do we communicate these qualities broadly to the appropriate audience?‖
2.3. Arts Pedagogy; What Can We Learn from Biomechanics Applications?
The use of biomechanics-based motor learning in sports pedagogy has proven highly
successful in improving human performance (Ballreich and Kuhlow, 1980; Magill, 2001). Its
focus has usually also been on enhancing physiological limits of athletes. Superficially, sport
and the performing arts may appear to have more or less in common, depending on the art.
For example, the athletic aspects of dance are obvious whereas those of music performance
often are not. All rely upon the training of human motor skills and the development of gross
and fine motor control. But, in music and dance the focus is not always overwhelmingly on
pushing physiological limits as it is in sports. Rather it is about controlling the process during
the performative act. If we continue with a sports analogy, in this regard arts performance can
be likened more to acrobatic kinds of sports than to those which rely mainly upon speed and
power. In arts performance, fine motor control seems to be a necessarily dominant concern.
Fundamentally, both sports and arts performance involve motor learning and the sublimation
of neural control to effect a desired physical outcome. In music, although success is judged by
an ―aural‖ outcome, it is a product of physical gestures. In dance, the result is a visual one
that is obviously guided by gesture. Since sports biomechanics is at an advanced stage of
application, it seems sensible to adopt and modify its already proven successful
methodologies to the context of the performing arts.
The perspective it can provide may contribute to a general demystification of the
teaching/learning process in the arts with a corresponding improvement in overall
pedagogical success as measured by both learning outcomes and injury prevention. When
adapting a sport-based motor-learning model to the arts, one must consider its
multidisciplinary character.
Figure 7. A comparison of the resources used in the pedagogical processes of sport and music.
In sports, intervention of related disciplines to influence pedagogy and learning is a

mature science; a team of specialists guides the learning/training. Interdisciplinary research is
commonly understood to hold significant potential for improving performance. In performing
arts education, responsibility rests primarily, if not solely, with the teacher who may only
have minimal training in some of the multiple required areas of competency (Figure 7). Even
in the area of arts medicine, prevention is an underutilized strategy for artists. Mostly,
medical attention is only sought for injury remediation.
The consequences of poor health and wellness practices for performers can be
devastating. They include pain, injury, long-term disability, and even career-ending
pathologies. With regard to the intervention of related disciplines, arts pedagogy is in a state
of development similar to that of sports pedagogy about fifty years ago. Learned behaviours
in the arts are dominated by the vision of the end result. In this regard it is fair to draw from
sport history (fifty years ago) as accurately representing and predicting the needs in the arts.
One sports anecdote that universally receives sympathy, acceptance and affirmation when
presented to the performing arts community involves Tom Simpson, British cyclist and world
professional road racing champion in 1965. His mind-set can be summarized by his two most
famous quotes. An advocate of performance enhancing drugs, he is noted as saying ―if ten
will kill me, give me nine‖. And his final words as he lay dying from cardiac arrest at the top
a mountain during a race were ―put me back on my bike‖ (Fotheringham, 2002). Arts
practitioners empathize with this ―pushing the limits‖ attitude.
Accordingly, the first aims of those who investigate Arts Biomechanics have been to
begin exploring the utility of methodologies and technologies already successfully employed
in mature fields of kinesiology and sports medicine to examine arts performance. The bases
for most of these explorations have been grounded in two fundamental practices of
kinesiology: motor control and motor learning. Not only do these provide a methodological
means to analyze movement, cognition and the learning process, they provide an improved
basis for discourse through an established commonality of idiom and terminology. Simply,
they provide a means for us to all speak the same language.
A Motor-Learning Model
Two foundational theories provide important perspective on skill acquisition in the
performing arts. The first, Ballreich‘s model for motor learning identifies three requisite
elements for success: (a) accurate characterization of the skill to be acquired, (b) skill transfer
between dissimilar bio-systems, and (c) skill acquisition without injury (Ballreich and
Baumann, 1996).
1. Characterization of the skill. Characterizing the skill to be acquired involves

scientific analysis to identify and describe motor control patterns (i.e. the interaction
and coordination of the neural and musculoskeletal systems). Methodologies from
kinesiology such as motion analysis are ideally suited to this purpose. Motor control
patterns common among expert performers can be generalized, establishing
―markers‖ which can be used as references in the pedagogical process. Using this
approach, characteristics of high-level performers can be contrasted with those
observed in learners. Such a quantification of fundamental skills begins to address
the most practical challenge in the arts – a clear understanding of motor control
patterns needed to communicate the desired outcome.
2. Skill transfer between dissimilar bio-systems. For a successful motor learning model,
fundamental skills must be understood in such a manner that the teacher can transfer
the principles to the student regardless of physical differences. The most significant
pedagogical weakness in ―experience-based‖ teaching approaches is the fact that not
everyone is the same anthropometric shape. Simply, the teacher‘s experience does
not really apply. Clear characterization of skills and marker-based learning can
facilitate skill-transfer between teacher and student by providing a general model of
the specific kinematic gestures vital to skill acquisition. However, for successful skill
transfer, ―universal‖ characterizations need to adjust for the ―particular‖,
compensating for anthropometric differences among learners. For this to occur, a
sufficient body of Arts Biomechanics research is required to inform teachers‘
practical experience.
3. Learning without injury. By directing attention to specific motions or neural control
patterns that novices must assimilate into their technique, skill transfer can occur
efficiently and effectively; reducing the assimilation time needed to acquire each
skill or technique, and thus reduce the risk of Repetitive Stress Injuries.
The second theory addresses skill development at the different stages of the learning
process. Relatively simple, the Fitts and Posner model has been proven effective in the
discipline of kinesiology (Magill, 2001). It identifies three general stages: cognitive,
associative and autonomous.
1. Cognitive. The beginner focuses on cognitively oriented problems. Performance is

highly variable, showing a lack of consistency from one attempt to the next. Often,
learners are confused about what they need to do in order to improve.
2. Associative. Cognitive activity changes because the learner has begun to associate
multiple references in order to guide the outcome. Through training and repetition
learners begin to make fewer gross errors as they refine and combine the multiple
skills acquired in the cognitive stage of the process.
3. Autonomous. Here the skill becomes almost automatic or habitual. Learners can
perform with repeated degrees of accuracy, without conscious thought.
Figure 8. The stages of motor learning adapted from models of Ballreich and Magill.
Both of these models address different aspects of motor learning. One deals with
biomechanical factors influencing the process of motor learning and the other concentrates on
the process of skill development (Ballreich and Baumann, 1996; Magill, 2001). Collectively,
these two models have important implications for the development of pedagogical practices in
the arts. Figure 8 is a combinatorial model, representing Ballreich‘s model for motor learning
on the vertical axis and Fitts and Posner‘s skill development model on the horizontal one
(Visentin et al., 2008). In our opinion, rationalizing this model for arts pedagogy requires
careful consideration and a nuanced understanding of disciplinary discourse within the arts.
One way in which application of the model may differ from a sports one lies in the emphasis
that must be placed at each stage of the learning process. Given arts pedagogy‘s tendency
toward teleology, new ideas need significant evidence (e.g. improved performance outcomes)
before they will be adopted. This requires: 1) clarity of definition, 2) convincing
methodology, 3) consistent means of situational comparison, and, 4) solutions that do not
compromise the artist‘s perceptions of the end product. Such a rationalization holds
significant potential for improving arts pedagogy.
This has clear ramifications for the respective roles and responsibilities of both teacher
and pupil. In the cognitive stage, the onus is on the teacher to clearly and accurately
characterize the skills to be acquired using instructional strategies such as modeling, clear
explanations, analogies, and a vocabulary that accurately describes the phenomena. In the
associative stage, teacher and student work more collaboratively toward the ―particular‖.
Compensation based on individual anthropometrics needs to be melded with aesthetic goals
during this stage, as external (e.g. biofeedback instructions from the teacher) and internal
processes (e.g. student‘s modifications to motor control according to the teacher‘s
instructions) are consciously assimilated. Multiple skills must be coordinated together at this
stage. In the autonomous stage of the process ―naturalization‖ occurs as skill control moves
from the conscious to the sub-conscious. Here, the onus for successful learning is mostly the
student‘s responsibility and the teacher‘s role becomes more focused on remediation. As skill
control becomes automatic, performers are freed to direct their attentions more toward
aesthetic, creative and artistic outcomes, which requires both forward planning and flexible
response to the circumstances of the moment. The quality of instruction and practice as well
as the amount of practice are important factors during all these developmental stages.
Arts Biomechanics has the potential to positively affect the performance arts through
influencing the pedagogical process. It builds upon already proven scientific models of motor
learning, adapting these to acknowledge artistic process and values. In the first two stages of
the learning process, it can directly influence the quality of teaching/learning that occurs. This
can accelerate the skill acquisition process and minimize the necessity of trial and error
practice. This will make the process a more enjoyable experience for the learner. This is a
motivational prerequisite since it frees a performer to transfer his/her attention from motor
control to artistic representation, indirectly improving artists‘ perceptions about the third stage
of the process.
Further, because arts performance involves long-term training, the dangers of
vocationally related injury due to poor pedagogical practice is very high. In this regard Arts
Biomechanics can highlight injury prevention as a fundamental ethic at each stage of the
motor development process. It can provide realistic and practical pedagogical strategies that
encourage changes in attitude and behavior. The benefits of this will be improvements to
artists‘ health, wellness and career longevity as well as enhanced self-perception of their
artistry.
2.4. The Tools of Arts Biomechanics Research
The scientific quantification of motor skills in arts performance is about decade old (G.
B. Shan and Visentin, 2003; Visentin and Shan, 2003). Prior to this, qualitative and some
quasi-quantitative image and video-based research exists in several formats; drawings,
photographic images, photographic time or multiple exposures, and video (Hilton, 1997;
Hodgson, 1958; Moyse, 1973; Palac, 1992; Tulchinsky and Riolo, 1994a). Existing research
has primarily been concerned with skill description as it relates to musical
values/interpretation and rudimentary analyses. Most of it can be found in medical,
pedagogical, and in cognitive/behavioral neuroscience (e.g. Music Perception) journals. As
such, biomechanics is mostly used as an indirect or supporting means to examine disciplinary
questions in these areas. As a primary area of study, Arts Biomechanics is still young.
Clearly, the technologies and methods of biomechanics are an attraction to researchers from
many disciplines, but the questions these researchers propose would likely be different than
those a biomechanist would proffer. In Arts biomechanics, the technology must be adapted to
the context of the arts. But, more importantly, the ―right‖ research questions need to be asked;
a condition that requires collaborative methodology. Since performing artists are intimately
involved with their art, objectivity can best be provided by an external observer. However,
since the artistic product is inherently dependent on the relationship between artist and
performance, analytic objectivity has functional limitations.
A review of the most recent literature reveals that research groups looking at the
performing arts are becoming more multidisciplinary. The reason for this is simple. Given the
complexity of the questions being asked, it is unrealistic to expect a single discipline to
provide the required artistic understanding, scientific expertise, and technological know-how.
Arts biomechanics as a new area of study can directly contribute in significant ways to the
latter two of these needs. Because it influences the health and wellness of performers, it
indirectly influences artistic outcomes as well. Biomechanics brings multiple tools and
scientific methodologies to bear on research which can be easily adapted to arts applications.
Currently, three technologies have dominated arts performance analysis: 3-D motion capture
(MoCap), force measurement, and electromyography (EMG). These technologies give rise to
additional tools, such as biomechanical modeling and multidimensional signal analysis of
simultaneous measurements (Visentin and Shan, 2004). The levels of sophistication currently
being demanded in human movement and performance research requires synchronized

information from multiple technologies in order to address more and more complex
questions.
Motion Capture
3-D motion capture can be used to track movement of any object. It does so by using
multiple cameras to view and triangulate the positions of markers that are placed on the item
(Figure 9). Some motion capture systems use passive reflective markers and others use active
optical markers (diodes). No matter the system, motion capture relies on multiple cameras
being able to see each marker since, for complex movements selected points can easily
become occluded by the measured object‘s orientation. At least two cameras must be able to
see each marker at all times for accurate reconstruction of XYZ coordinate data.
Using sophisticated software, precision motion capture measurements can be assembled
in 3-D computer space, supplying kinematics (positional data, distances, velocities, and
accelerations) of the movement. For human subjects, markers need to identify functional
anatomical points, such as joints (Lundberg, 1996). Typically a special garment that
effectively eliminates the relative movement of clothing and skin, but without restricting
motion, is worn by subjects; markers are appropriately positioned on this garment. For
example, upper body movement can be quantitatively characterized using markers placed on
trunk and joints as follows: the Sternal end of Clavicle, Xiphoid process of Sternum, 7th
cervical and 10th thorax vertibra, for the trunk; right and left Acromion, lateral Epicondyle,
and Styloid processes of Radius and Ulna for the arms, and; 3rd metacarpal and finger joints
for the hands. Head orientation can be determined by placing markers on the head. Frequently
additional reference markers are used to determine segmental rotation. If any equipment is
used during the activity, its movements can be captured by placing markers on the items. In
all instances at least three markers are needed for each measured body segment and each
piece of equipment in order to determine both positional translation and rotation.
Figure 9. 3-D motion caption of Grande Jeté (Ballet): the set-up of the 9 cameras and a sample frame
(G. Shan, 2005).
Biomechanical Modeling
From changes in position over time, computer analysis takes simple kinematic data and
translates it into skeletal and muscular movement using the fundamental precepts of physics
and anatomy. This is biomechanical modeling. In such biomechanical modeling, inertial
characteristics of the body are estimated using anthropometric ―norms‖ found through
statistical studies (Shan and Bohn, 2003). Effectively, biomechanical modeling can serve as a
platform for data analysis, providing the means to integrate information collected multiple
measuring systems such as motion capture, force measurements and EMG.
Task-specific biomechanical models are used for each activity. Modeling techniques such
as dynamic and inverse dynamic analysis allow researchers to establish joint coordination
(motor control patterns), internal loads (moments and forces) and muscle working patterns
(loading types). This information is used to study interactions between humans and their tools
(e.g. a violinist and his bow or a golfer and her club), the interface of the body with its
surroundings (e.g. dancing) muscle group timing/coordination (e.g. multi-joint tasks such as
piano playing), and neural control (Timing article-peter). For example, studies by Shan and
Visentin (2003 and 2004) analyzing violin performance used 30 reflective markers to track
professional violinists‘ left and right-arm skills. Cameras were placed in a circle around the
subjects, creating a capture volume of approximately 3m  3m  2.5m. Of the thirty markers,
24 were placed on the upper body of each subject and 3 were positioned on each the violin
and bow. Marker trajectories were used to reconstruct a biomechanical model of violin
performance. (Figure 10).
Figure 10. The set-up of 3D motion capture system and the 10-segment biomechanical model of violin
playing.
Electromyography
Electromyography is a seductive technology because it provides researchers easy access
to data about the physiological processes that cause muscles to generate force, produce
movement and engage in the countless functions that allow humans to interact with their
world. By placing electrodes on the surface of the skin directly above the longitudinal and
medial-lateral centre of a muscle, minute electrical activity can be detected during muscle
contractions. These seizure potentials can be directly correlated with the intensity and
duration of muscle activity. As such, EMG supplies researchers with a means to determine the
timing of neural control and to fractionate the response process of coordination among muscle
groups.
EMGs main attraction is that we can to some extent examine what happens to muscles
inside the body using non-invasive means. In absolute terms, seizure potentials measured by
EMG are very small and an amplifier must be used to enhance electrical signals, which
typically register under 10 millivolts. Further, the amplitude of the signal is dependent on the
physical characteristics of the subject (e.g. skin conductivity, thickness of fat layers between
muscle and skin, training-dependent neural control, etc.). So, comparison of raw data among
subjects is senseless. Most significantly EMG signals are stochastic (random) in nature. These
would seem to be significant challenges to EMGs utility. But, processing provides a means to
overcome such barriers.
In order to interpret EMG data, which is a relative (subject dependent) rather than an
absolute value (subject independent), a variety of processing protocols exist. Enveloping is
the one most commonly found in the literature (Winter, 1984; Nandedkar and Sanders, 1990;
Shiavi, Frigo and Pedotti, 1998). The enveloping process involves full-wave rectification of
raw EMG signals and low-pass filtering (Figure 11). Such filtering improves EMG sensitivity
as a diagnostic tool by keeping dominant components of the signal and eliminating unstable
ones. According to the literature, a 6 Hz filter Fast Fourier Transformation (FFT) low-pass
filter is typical for human activities (Arsenault, Winter and Marteniuk, 1986; Shiavi et
al.1998).
Enveloping provides a stable signal for researchers. However, it must be noted that such
signals are non-quantitative measures of neural-musclar activity. Signal intensity cannot be
used as a comparative value between subjects, or even between trials of the same subject held
on different days, since it is impossible to exactly replicate placement of electrodes and other
conditions that influence the quality of EMG signals. Hence in order to provide quasi-
quantification and possibility of comparison between trials, a normalization process is
required. Normalization expresses all measurements as a percentage of a chosen reference
value. For EMG, commonly used references include: 1) the maximum signal level during a
trial, or 2) a Maximum Voluntary Contraction (MVC) determined in a separate trial under
high-load conditions. Using the first of these provides results of questionable quantitative
utility since the ―reference‖ may vary from trial to trial. The second approach establishes the
physiological limits of the muscle under static contraction, a value which will not change
from trial to trial. This allows researchers to examine muscle load intensity in percentage
terms of the subject‘s MVC (or physiological limits). Such a value can provide a reliable
means of comparison among different subjects and trials.
Figure 11. EMG of rectus abdominis during voluntary breathing – the process establishing enveloped
and normalized EMG for quasi-quantification.
Figure 12. Rectus abdominis activity at different pitches.
Figure 12 is a composite of three different trials measuring the rectus abdominis activity
of a professional opera singer (mezzo-soprano). The rectus abdominis MVC was determined
in a separate trial under high loading conditions. The protocol involved two sets of singing
four short onsets (intonation of a pitch) followed by a sustained tone produced on the same
pitch.
There was approximately three seconds of rest between the sets. The same protocol was
used at three different pitch frequencies – high (black), medium (dark grey) and low (light
grey) for a typical mezzo‘s vocal range.
In terms of vocal production, analysis of Figure 12 supplies rich details. Normalized
EMG measurements are particularly revealing and the consistency/repeatability between sets
provides evidence of high levels of training. Clearly, different pitch levels required
significantly different amounts of physical effort from the singer. At low and medium pitch,
there was almost no noticeable difference in physical effort between short onsets and long
sustained tones. Interestingly, for low and medium pitch there was almost no difference in
rectus abdominis activity during periods of singing or rest. Physical effort remained under
20% of the subject‘s MVC.
At high pitch, the story changes significantly. Peak levels of EMG were above 60% of
the MVC. This represents a physically demanding task. For the four short onsets, a clear
pattern existed – first onset lower and remaining three higher – one that may be explained in
terms of the individual‘s technique. Notably, the period of rest for the high frequency trial,
showed lower values than the periods of rest for the other two ―easier‖ ones (low and
medium).
Further, the part of the trial where a tone was sustained began with an onset of about 40%
of MVC and gradually increased to peak levels of the short onsets as the breath of the singer
was expended. These two facts suggest a subconscious pre-planning on the part of the singer,
since experience would tell her that an increase of physical effort was going to be required
during the generation of a sustained tone. This has implications for vocal pedagogy, since the
singer was not perceptibly aware that such patterns were developing. Clearly, in spite of its
quantitative limitations, EMG can provide significant information in terms of skill analysis
and evaluation of training.
Force Measurements
Force measurements provide quantitative measures of the interaction between a human
body and its surroundings. They may be determined using platforms, small sensors, arrays of
sensors, or in a myriad of other ways specifically required by the conditions of testing.
In arts biomechanics, force platforms can be used to quantify ground reaction forces and
weight transfer during activities such as dancing or balancing. Figure 13 shows a ballet
pirouette.
The yellow arrow shows the force vector of the ground reaction force in relation to the
dancer‘s inertial centre of gravity. Not only can such information be used for pedagogical
purposes, it has potential industrial significance for design of ballet footwear and for injury
prevention and risk management regimes. (Chockley, 2008; Shippen and May, 2008; Shippen
and May, 2009; Solomon and Solomon, 2004)
Small sensors in a variety of configurations can be used to answer task-specific research
questions. Examples in arts biomechanics includes installation in keyboards (Parlitz, Peschel,
and Altenmuller, 1998), in mouthpieces of wind instruments (Devroop and Chesky, 2002),
and on places where performers fingers or thumb come into contact with an instrument
(Nemoto and Arino, 2007). The data gathered gives information about finger pain, efficiency
and effectiveness of physical effort and the effects of long-term training.
Figure 13. Measurement of ground reaction force using platform during dancing (Shippen, 2009).
Synchronized Measurements
Together, data collection systems such as the ones described supply information on how
individuals interact with their environments and tools. The data delineates cognitive and
neuromuscular processes used while preparing to produce and during the execution of an
action. Synchronized systems – those where data from multiple systems are measured and
time-coded simultaneously – are the best means to analyze and understand the complexities
of human performance because they provide multiple perspectives of the same phenomenon.
The real challenge lies in find meaningful ways to integrate these various perspectives into a
holistic view. From an analytic standpoint, it is problematic that inputs can include many
different time variables (energy, force, linear and angular positions, speeds, accelerations,
bio-signals), each of which may be measured in different units, sampling frequencies and
scales. Thus, for meaningful quantitative research, a central analytical tool needs to serve as a
nexus for comparing and correlating the multiple signals that are generated. Biomechanical
modeling provides just this means (Figure 14).
In summary, synchronized systems coupled with biomechanical modelling reveal more
anatomical and biomechanical perspectives than measurements alone. 3D motion capture
system supplies the spatial position, velocity, and acceleration of limbs and/or joints. EMG
captures muscle activity, which can be utilized to fractionate the response process and get
intensity and timing of neural control as well as the coordination between and among muscle
groups. Force measurements reveal the interactions between subject and environment. Video
other means provide a traditional external view for motion analysis reference. Biomechanical
modeling links the coordination of the joints with neural controls (muscle activities). In total,
synchronized systems present a more global technique to measure, analyze and understand
any human motor skills.
Figure 14. Biomechanical modeling as a platform for integration. Modeling supplies the means to
connect and compare multiple inputs. Single-headed arrows represent inputs that are measured. Double
arrow heads represent those that both influence and are influenced by the model (Visentin and Shan,
2004).
2.5. Samples of Pedagogic Applications
Section 2.3 began with a discussion of the principal challenges faced by Arts
Biomechanics as a field of study and ended by identifying two questions that Arts
Biomechanics can begin to answer. The first of these, ―Can we identify, describe, and
quantify qualities that lead to artistic success while minimizing the potential of performance
injury?‖ directly addresses pedagogy. Its goal of reducing the potential of performance
injuries will ultimately be unsuccessful unless it provides information for improving
pedagogy. Both teachers and learners are primarily interested in much more primitive
questions, such as ―How can I teach/learn skills faster?‖ Concerns for injuries typically come
low on both students‘ and teachers‘ priority lists until they have actually occurred. By
engaging in research that addresses both effectiveness and efficiencies in performance, Arts
Biomechanics can begin to influence attitudes. If performing artists believe there is a
faster/better way to do something, they will listen. But, research in Arts Biomechanics will
have to provide practical and convincing scientific evidence, clear language and definitions,
and a means to re-evaluate and improve pedagogy in ways that do not compromise artists‘
perceptions of the end product – their art.
Unfortunately there are relatively few examples of quantitative research in the literature.
One example, related to violin, gives a glimpse of the potential of Arts Biomechanics
applications in pedagogy (Visentin et al., 2008). In the study the authors set out to determine
whether or not it was possible to accurately and objectively characterize and generalize some
biomechanical skills related to violin performance. Both smooth and bounced bowing
techniques were examined at various speeds and on all four strings of the violin. The goal was
to accurately identify postural and kinematic landmarks, kinetic ―feel‖, and motor control
patterns (i.e. ranges of motion of single joints and dynamic interactions among joints) that
might characterize the movements of expert performers. From a pedagogical standpoint,
consistencies observed among expert subjects of differing anthropometric proportions can

provide ―markers‖ that should be easily employable in teaching.
Figure 15. Right elbow and shoulder flexion/extension. Ranges of motion (ROMs) are measured by
peak to trough differentials (Visentin, Shan and Wasiack, 2008).
The results of the study showed that, for the two identified bowing techniques, finer
motor control movements, such as those found in the wrists, were highly variable among
subjects. However, shoulder and elbow movements, those joints that dominated gross motor
control, provided generalizable patterns. Using these, several pedagogical markers were
identified and quantitatively described. These markers provide information about posture,
positions, and time-based dynamic interactions among joints during performance.
Figure 15 shows both that there were absolute changes in the amplitudes of the joints
measured and that the control pattern changes of the two joints appeared to be complimentary
– one ROM gets bigger as the other gets smaller. From G string to E string overall shoulder
flexion/extension angle decreased and shoulder flexion became more active and the ROM
increased. The elbow acted in exactly the opposite manner. This is one example of
establishing a compound ―marker‖ for a multi-joint task; one that can be used during the
teaching/learning process.
Additionally, through biomechanical modeling the study identified tempo-dependent
phases of motor control. Three discrete control pattern phases were found for both smooth
and bounced bowing (G. B. Shan et al., 2007; Visentin et al., 2008). Forward dynamic
analysis revealed each phase to have distinct combination of kinematic factors. The first
phase was marked by increased bow acceleration and speed (indicators of exertion), while
travel distance of the bow decreased (reaction showing a compensation). The second phase
was characterized by constant bow acceleration, revealing workload to remain steady, while
distance and speed decreased (showing efficiency gains). And the third, revealing an
approach to physical limits of the player, showed bow acceleration, speed and distance to all
gradually approach their respective plateaus (Figure 16).
Figure 16. Kinematic factors showing three discrete phases of motor control defined by two critical
points (vertical dotted lines). From left to right: increasing physical effort, optimization, and
approaching physiological limits (Shan and Visentin, 2003).
Although there are relatively few examples in the pedagogy literature, other sources
contain information that also has pedagogic implications. The problem with this is that the
venue of the information might preclude reaching those who are most influential in the artistic
practice – performers and teachers. One such example comes from the medical literature (G.
B. Shan and Visentin, 2003). The study was directed toward quantifying biomechanical
factors related to Repetitive Stress Injuries (RSIs). Findings suggested that music education
would benefit by applying biomechanics in the teaching and learning process. For example,
Figure 17 represents a trial typically found among professional violinists. The elbow heights
of the subject were normalized as a percentage of his body height.
Figure 17. Vertical movement of left and right elbows during a typical 32-second trial. A two-octave G
major scale was performed, beginning on the G-string cycling to E-string and back (G. B. Shan and
Visentin, 2003).
This kind of kinematic analysis shows asymmetrical control between left and right side of
the body. Elbow heights were near static on the left (approximately 68% BH), while the right
showed a complex dynamism. Right elbow ROMs were remarkably greater on the E-string
(9.3% of body height) than on the G-string (6.9% of body height). The implications for
pedagogy are significant. When teaching about bowing, the teacher‘s efforts need to convey
the dynamic nature of the right elbow; its position depending on both string played and part
of bow used. Further, teachers need to be aware of how the kinematic chain actually works.
The mechanism that raises and lowers the elbow is actually the shoulder. Failure to clearly
express such markers can slow the learning process, limit the student‘s potential, or be a
fundamental cause of eventual injuries.
A marker-based teaching approach can provide insight into how it actually feels to play
the violin. As these studies suggested, such methodology may have significant implications
for development of new pedagogy models for learning performance skills on the violin.
Connecting objective data with subjective perception or ―feel‖ is one way to accelerate the
motor learning process. Teachers and students alike want to reduce the time required for skill
acquisition and an evidence based approach decreases the time students must spend
―experimenting‖ to establish basic coordination control. Clearly, more studies are needed to
develop such models, not only for the violin, but for other instruments, for dance and for the
related arts. A recent study examining university level trumpet players also alludes to the
value of quantitative analyses for the development of new approaches to pedagogy (Devroop
and Chesky, 2002).
The selected examples show Biomechanics to be ideally suited for such task. Marker-
based approaches to performance pedagogy have the potential to increase the efficiency and
effectiveness of skill acquisition. In this way, Arts Biomechanics is a new application of well-
established ideas. These ideas can guide the formative stages of learning and influence arts
professionals and teachers. But, to be successful, it needs to meet the ―Art‖ on its own terms.
To do so, Arts Biomechanics must be willing to engage in a deeper discourse that explores
the relationship between motor-control patterns (the ―quantitative‖) and the subjective aspects
of the art – the desired aesthetic outcomes; thereby preserving natural variability and
individuality of artistic ―style‖.
Future Directions
Arts Biomechanics is a field in its infancy. In can challenge traditional practices in arts
pedagogy by successfully employing quantitative techniques to objectively describe
physiological and biomechanical phenomena associated with fundamental performance skills.
Systematic scientific inquiries hold the potential to provide answers to important multifaceted
question such as those posed at the beginning of this section. However, arts performance is
different from most day-to-day human motor activities. All arts performances are influenced
by a complex set of factors related to the performers‘ personal aesthetic sensibilities and
perceptions. Quantitative analyses can challenge these self-perceptions. The studies identified
in this chapter show ways to establish a science-based approach to pedagogy. Such
methodologies can increase the efficiency and effectiveness of skill acquisition, which will be
welcomed by the arts industry – but only if they provide practical and convincing scientific
evidence, clear definitions, and a means to reevaluate and improve pedagogy in ways that do
not compromise artists‘ perceptions of the end product. It remains to explore the usefulness of
Arts Biomechanics in the study of multiple Arts. A more comprehensive body of research
could potentially revolutionize how performance skills are taught.
3. INJURY RISK QUANTIFICATION AND PREVENTION RELATED TO

ARTISTIC PERFORMANCE
3.1. Overuse Syndrome
Overuse syndrome is a ―catch-all‖ term for many different types of repetitive stress
injuries (RSIs). Among performing artists, a myriad of injuries occur. Because artistic success
or failure is often determined by micro-increments of ability, the smallest injury can
negatively affect the performer and the artistic outcome. For example, an injury that affects
the range of motion of someone‘s big toe by 10% would go unnoticed by an office clerk.
However, for a prima ballerina, it could be the beginning of the end of a career. Pathologies
can include: tennis elbow, shoulder injuries, rotator cuff injuries, focal dystonias, tendonitis
and lower back pain and more (Fry, 1987; Zaza, 1992, 1998). Although repetitive movement
may typically be the mechanism of these injuries, causal factors are deeply rooted in the Arts
veneration of their traditions, pedagogic and artistic.
Reported rates of vocational injuries among performing artists are very high. Among
musicians, affect between 50% and 76% of professionals (Fry, 1988; Fry et al., 1988; Zaza,
1998). Considering that this applies to musicians with injuries serious enough to affect their
vocational activity, often for extended periods and sometime career-ending, these statistics
are staggering. There are three general categories of injury commonly diagnosed and RSIs
accounts for over half of these. Given the high incidence of RSIs, the industry has recognized
the need for developing efficient and effective prevention and risk-management strategies.
Currently most efforts in this regard are still in the early stages of research. Biomechanics has
been contributing in the areas of accurate and quantitative description, understanding the
nature of the physiological loading, and linking epidemiological studies to the observed
activities. The aim has been to minimize or prevent injuries by reducing biological loading.
Since Arts performance activities often involve the whole body, and demand high degrees of
both fine and gross motor control, an examination of their complex biomechanical processes
may lead to generalizations about RSIs that apply to other complex human activities.
Definitions influence research questions, design and structure. Therefore, it is necessary
to review traditionally held definitions in order to ensure that they are sufficiently nuanced to
direct researchers in the design of their studies. RSI has been traditionally defined as the
excessive use of body parts so that the accumulation of micro- trauma exceeds human
physiologic limits (D. M. Dawson, Hallett, and Millender, 1990). It has been understood for
quite some time that basic factors influencing the etiology of RSIs include: (1) genetic
makeup of indificuals; (2) performing technique; and (3) time  intensity of activity (Fry,
1987). Recent studies, identified below, improve upon these basic definitions.
3.2. Biomechanical Modeling as a Platform to Understand OS
The details of this section are drawn from an article published in the journal Medical
Problems of Performing Artists (G. B. Shan, Visentin, and Schultz, 2004).
Injuries that occur from repetitive use significantly differ from those that typically occur
from load-related accidents or during sports. In the case of the latter, injuries are essentially
due to load impacts suddenly exceeding human physiological limits - limits that are
theoretically possible to evaluate, and are relatively easy to diagnose. On the other hand,
activities that cause RSIs have characteristic loads well below human physiological limits and
injuries result from multiple micro-trauma, the effects of which accumulate over time. A
gradual and imperceptible onset of injury shows RSIs to challenge human physiological
tolerance - a factor that is difficult, if not impossible to quantify. Even though it may be
tacitly understood by those who actually deal with such injuries, some current definitions still
fail to articulate this distinction. Using the term ―physiological limits‖ in connection with
―micro-trauma accumulation‖ is misleading. RSIs are entirely different than impact injuries
because of the time-based and cumulative nature of the pathology. Thus, for meaningful
quantitative research, a central analytical tool needs to serve as a nexus for correlating the
multiple signals that are generated and for situational comparison.
In the case of sport injuries, biomechanical modeling has successfully served as this
nexus for assessment, remediation and prevention purposes (Ballreich and Baumann, 1996;
G. B. Shan, 1999). It seems the next logical step to employ biomechanical modeling as a
platform in the service of RSI research. However, generalized models must first be proven to
be theoretically valid for the human body, and then modified to accommodate the conditions
that accompany RSIs. Ultimately, it is hoped that this will lead to research that can clearly
explore the concept of physiological tolerance where: 1) a temporal factor (the process of
micro-trauma accumulation) is evaluated, 2) rates of recovery during periods of relative
inactivity or rest are considered, and 3) increased understanding of the evolutionary nature of
OS injuries will result in changed attitudes in the workplace and a focus on the development
of prevention strategies.
Figure 18. Biomechanical model for quantification of musculo-skeletal loading and muscle
lengthening/shortening.
As it is the mechanical work performed by OS patients that obviously causes observed

damage, an evaluation of ―work‖ appears to be a logical first step. Biomechanical modeling
permits quantification of both the characteristics of the human body during activity and of the
workload levels (G. B. Shan and Visentin, 2003; Visentin and Shan, 2003). It is the tool best
suited to this task and can be used to make connections between the data generated by
multiple assessment techniques (Figure 18). Simply, biomechanical modeling allows us to
understand what happens inside the body using information gathered from outside the body.
3.3. The Theory of Biomechanical Modeling as It Applies to OS
The following discussion is based on several biomechanical studies (G. B. Shan et al.,
2007; Visentin et al., 2008; Visentin and Shan, 2003, 2004). Using motion capture
technology, activities can be analyzed using biomechanical modeling. Anatomical positions
can be obtained, the skeletal structure and simple positional data can be translated into
skeletal movement using the fundamental precepts of physics. In such modeling, inertial
characteristics of the body are estimated using anthropometric ―norms‖ found through
statistical studies (G. B. Shan and Bohn, 2003; Winter, 1990). Combined with motion capture
information, these allow the calculation of load intensity (which can be described using a
variety of parameters measured at the periphery of the body) resulting in factors such as:
force, work, moment, (angular) velocity or acceleration and Head Injury Code index (HIC)
(C. M. Gadd, 1966; Hennig and Lafortune, 1991; Kosiak, 1959). Biomechanical modeling
allows muscles to be added onto the skeletal frame using anatomical knowledge regarding
normal attachment points. Thus, muscle lengthening and shortening can be determined in
connection with skeletal movement. From these, 1) internal loads (bone-to-bone contact force
and muscle force, rotation caused by joint moment) can be quantified and, 2) when correlated
with electromyographic signals, muscular work patterns (eccentric, concentric, isometric) can
be determined.
All ―modeling‖ involves simplifications of reality. The method‘s strength is that it
provides a means to understand complex systems without becoming lost in detail. In the case
of the human body, the key justification for using biomechanical modeling as an analytical
tool hinges on the accuracy of internal load calculations based on the fundamental precepts of
Newtonian physics (classical mechanics). To establish the contextual validity of the
calculations, several steps need to be taken:
a) the various forms of mechanical work (W) as they pertain to classical mechanics
must be established
b) a generalized form must be devised
c) the generalization must be justifiable and applicable to all cases for biological
reactions in the human body
d) the system must be related to RSI specific conditions.
3.3.1. The Physics of Mechanical Work

Depending on the situation, work (W) may be mathematically represented as:
 Work of force F over distance d W=Fd

 Work of force F at velocity v over time period t: W=Fvt
 Work of moment M at lever arm length l: W=M/ld
 Work for accelerating (a) a mass m from 0 to velocity v: W=½mv²
 Work for accelerating a mass m to momentum p: W=½p²/m
 Work for accelerating mass m by force F over time period t: W=½(Ft)²/m
 Work against friction
 (c is a coefficient, n depends on the kind of friction): W=cvnd
Note: for the sake of simplicity, all vectors are considered constant and parallel or
orthogonal to each other.
3.3.2. The Generalized Form Related to Work

In order to overcome the diversity of form observed above, it is possible to establish a
generalized form covering many (or possibly all) cases through the following steps:
By appropriately substituting d = vt and v = at and p = mv, these expressions can be
expressed in alternate forms as follows:
W=Fd becomes W=Fvt

W=Fvt remains W=Fvt
W=M/ld becomes W=(M/l)vt
W=½mv² becomes W=½mavt
W=½p²/m becomes W=½mavt

W=½(Ft)²/m becomes W=½Fvt
W=cvnd becomes W=cvnvt
In summary, unlike variables such as force, which cannot in this case be practicably
considered a parameter due to its time-dependent nature, W may be used as a parameter,
establishing a single value for a given process. In each of the above:
 there is a case specific Variable of Intensity (VI) - which may include force, moment,
stress, velocity, acceleration – plus transforming variables
 they are weighted differently by using an activity-specific exponential factor n
 loading duration t is taken into account (which may also have an exponential factor,
m).
Thus, the final mathematical representation in all cases is:
W=CVI n t) m
where C is a coefficient constant, n and m are integers and VI is the case-specific variable of
intensity.
It should be noted that, in addition to Work, three other parameters are also frequently
used for predicting biological reactions: momentum (P), load duration (t), and maximum
stress (MS). For demonstrating a common structure, they can be written in unusual ways
using force as a common factor. Substituting these:
P=F 1 t
t= F 0 t.
MS= cF(t) n dt.
NOTE: In this last equation, t he maximum force determines the value of the integral.
With high n, injury occurs as t0 and mechanical load approaches or exceeds physiological
limits (Hennig and Lafortune, 1991). This is the condition of impact injuries.
From the above it can be seen that all have the structure:
CF(t) n dt) m
when W, P and t are written as integrals.
3.3.3. Biomechanics and Biological Reactions in the Human Body

As the above are purely mechanical, it may be questionable that they really describe the
reaction of a biological system. To demonstrate this, static work is most cited for showing the
difference between mechanical and biological systems in terms of loading. When holding a
mass, mechanical work is zero; physiological work is not. But this is an inadequate
illustration as it confuses the techniques of ―arresting‖ and ―active holding‖. For example:
when standing with bent knees physiologic work is needed, but when standing with
straightened (arrested) knees, work is much less.
In consideration of this, empirical in-depth biomechanics studies yield parameters of
biological reaction (PBRs), which are remarkably different from those yielded by mechanical
ones. Examples from the literature include:
 The Gadd Severity Index a 2 . 5 dt= const, predecessor of the similarly defined Head
Injury Criterion HIC is widely used for predicting skull damage, where a certain
constant is believed to be related to a certain degree of damage (Administration,
1996; Barth, 2000; C. Gadd, 1966; McLean and Anderson, 1997; USA, ;
USGovernment).
 Re-evaluation of data indicating the p/T-limit where ulcers were seen when pressure
p was exerted to the skin of dogs over time T revealed that the best PBR describing
the limit is p 1 . 2 const. (Kosiak, 1959).
 Load which causes increase of bone density is best described by a n dt, where n
varies between 4 and 12, depending on the situation. (Nicol, 1998).
 World records in running are described with very high precision by v n dt=const.
The distance range studied was 12.5m to beyond a marathon and could be divided
into four ranges with highly constant n in each range. Between ranges, the n varied
from 0.006 to 0.229. As running a world record describes the maximum load that can
be bourn in this situation. This formula is a PBR too. (Nicol, 1996, 1998)
In spite of the differences, it can be seen that these studies result in PBRs which have a
structure in common with the mechanic ones:
VI n dt) m =constant
The differences are that:
 in mechanical PBRs, m may be an integer larger than zero, whereas biomechanical

PBR require m=1
 in the case of biomechanical PBRs, n does not have to be an integer because the
empirical data may show a ―best fit‖ to require a non-integer.
 besides PBRs that can be written as integrals over the whole duration of VI,
biomechanic PBRs are used which relate to only a part of the activity duration. The
criterion of Hennig and Lafortune (1982) is an example. It looks only at the range
between the onset and the maximum of the VI.
3.3.4. Characteristics of the PBR related to OS

As it is the mechanical work performed by OS patients that obviously causes observed
damage, a work-related model of injury risk has validity. For OS the PBR can be made
situation (activity) specific by examining internal and external loads and integrating their
influence into risk assessment. Injury risk predictions may thus be a product of this PBR.
Whereas impact injury theory examines loads that exceed physiological limits, OS occurs
under different circumstances and requires an investigation of physiological tolerance.
The following simple model of OS is considered:
 OS is the result of the accumulation of micro-trauma

 activity causes micro-trauma damage over time
 biological repair takes place during periods of relative inactivity
 the cycle of 2 and 3 above is repeated multiple times
 with the result that the micro damage accumulates because the repair cycles are
insufficient for complete recovery
 finally, a critical limit is exceeded, macroscopic injury may be diagnosed, and the
physiological tolerance of the individual is surpassed.
In quantitative terms, the following fundamental premises may be used:
1. Since damage (D) = MicroDamage  , it is obvious to put damage

proportional to time T, because microdamage may be considered a constant that
depends on the loading situation and on the properties of the biological system (both
of which change minimally).
2. As repair (R) = R  T. The repair parameter R may be considered constant too.
(As an alternative, R can increase or decrease as total damage changes due to the
ability of the biological system to exercise repair under different conditions – i.e. the
body may become more or less capable of repair at different levels of injury.
Considering R to be constant represents a midpoint between these alternatives.).
3. Cumulative Damage (CD) is thus the difference between damage and repair.
CD=DT  RT.
Figure 9 shows the development of CD over a single cycle where slow repair over a long
time leads to complete repair (blue lines), a single cycle where fast repair over a long time
leads to complete repair (green) and repeated loading cycles (red) where CD finally exceeds
the critical CD value (orange).
From this and other current studies (Visentin et al., 2008; Visentin and Shan, 2004), the
following hypotheses may be made:
1. that loads leading to Overuse Syndrome (OS) have qualitative and quantitative
factors, both of which must be considered when undertaking risk assessment and
developing prevention strategies
2. that qualitative factors are related to the type of load (impact, fundamental, static,
quasi-static, or dynamic) and the quantitative factor is related to the magnitude and
duration aspects of the loading
3. that, at comparable load levels, static or quasi-static loads are more injurious than
dynamic ones; making quality of load a primary factor to consider when assessing
injury risk
4. that the intensity of a series of dynamic load cycles with small intermediate recovery
phases may be considerably higher than the intensity of static load of equal duration
without causing injury
5. that risk is directly related to the PBR and Intensityn  Duration of Activity, and
6. that Intensity and n are influenced by quality and quantity of load.
Damage (slow) Repair (slow) Cummulative damage (slow)

Damage (fast) Repair (fast) Cummulative damage (fast)
Damage (repeated) Repair (repeated) Cummulative damage (repeated)
Critical cummulative damage
Damage, Repair
Time
Figure 19. Cumulative damage scenarios from cycles of micro-damage and repair (Visentin et al, 2004).
In quantitative time-dependent terms, the PBR is directly related to activity Intensity (I),
which is contingent on the patterns of the load for the activity in question. Therefore:
PBR  (I)ndt.
For RSIs, the intensity of activity is very low, work patterns change over time, and
duration is the prime factor. Such conditions make RSI research extremely complex since
establishing injury risk requires that: 1) intensity (I) be measured for the activity; 2) duration
of activity be integrated into the risk calculation, resulting in a context-specific PBR; and 3)
that rates of repair during periods of rest be considered. Unlike impact injuries where there is
virtually no time between loading and injury (t0), repair cycles during periods of relative
inactivity may not be neglected in the case of OS. In representing this model mathematically,
the repair parameter (R) must be considered. Thus, Accumulated Micro-Trauma (AMT) is the
net of these two biological processes:
AMT = (I)ndt  Rdt
All research must be based on some fundamental theory. This work establishes a more
nuanced general form to guide RSI research in the future.
3.4. Current OS Studies Related to Music Performance
Systematic research into artists‘ vocational injuries is relatively young. The primary
venue for knowledge transfer has been the journal Medical Problems of Performing Artists. It
has a 27-year history examining epidemiology and remedial treatments for artists‘ injuries. It
has long been acknowledged that prevention was an important, but under-addressed area. To
promote research in prevention, the journal has been very accepting of cross-disciplinary
research that provides vocational views and/or points in the direction of remediation.
Quantitative biomechanics studies have been an important part of providing linkages between
epidemiology, remediation and prevention since 2003. The earliest of these studies examines
biomechanics of violin performance (G. B. Shan and Visentin, 2003; Visentin and Shan,
2003). Prior to this, some qualitative biomechanics work was included and, although it
implicitly showed the potential to influence prevention, the results of the studies were
essentially just descriptive (Tulchinsky, 1995; Tulchinsky and Riolo, 1994b). One of the
missing elements was a theory or framework guiding future steps in the process.
Richard Dawson, one of the founders of Medical Problems of Performing Artists, has
published two retrospectives documenting the kinds and numbers of studies found in the
international literature pertaining specifically to artists‘ injuries (W. J. Dawson, 2003, 2007).
These provide insight into the state of research involving biomechanics and the Arts. Of 2385
studies identified by Dawson (between 1997 and 2006), only 23 involved biomechanics. The
vast majority of these used the techniques and technologies of biomechanics to examine
causal factors leading to repetitive stress injuries. Below are some examples of Arts
Biomechanics research in the literature:
The body of work by Shan and Visentin has been formative in establishing a sequential
series of steps and a fundamental theory underpinning the utility of Arts Biomechanics in RSI
research (G. B. Shan and Visentin, 2003; G. B. Shan et al., 2004; G. B. Shan et al., 2007;
Visentin et al., 2008; Visentin and Shan, 2003, 2004). From phenomenological description to
multidimensional signal analysis, Arts Biomechanics was shown to hold significant potential
in the areas of loading quantification and injury risk management. Equally important, it was
shown to have potential in the areas of changing vocational attitudes and pedagogy. One of
these studies identified three factors influencing the relationship between muscle loading and
RSIs: load quantity (intensity), quality (static versus dynamic), and duration (time) (Kuipers,
1998; Taylor, 1996; Visentin and Shan, 2004). Regarding load quantity, the study showed
muscle activity during violin performance to be within the ―safe‖ range as typically identified
by the sports biomechanics literature (Peterson and Renström, 1987). Given the frequency of
injury among violinists this is obviously not the case. Shan and Visentin‘s work also
considered load quality and duration. Using multidimensional signal analysis, loading
patterns (qualities) in the muscles and joints were correlated with both load quantities and
injury rates. Because violin performance is an asymmetrical activity, this provided a means of
comparison between static and dynamic loading. Results had implications for the relative
risks of static vs dynamic loading and began to quantify relative levels of risk. They also
pointed toward pedagogical strategies aimed at prevention. Generally speaking, since static
loading is more dangerous than dynamic (for the etiology of RSIs), the study suggested
means to minimize static conditions or postures, and encourage dynamic motor control.
Further the body of work addressed the issue of disconnects between performers bodily self-
awareness during performance and the quantitative reality of the phenomenon. For example,
violinists are typically taught that the left elbow should have a degree of lateral movement to
accommodate playing on different strings. This improves performance outcomes in the areas
of tone production and pitch tuning. Unexpectedly, they found that performers generally
locked their shoulders in a highly adducted position; a posture that creates static conditions
and increases the risk of injury to the left shoulder (Visentin and Shan, 2003).
In another violin study by Rabuffetti et al. (2007), the effects of using different violin
shoulder rests during performance were quantitatively described. The positions of markers
placed on the player‘s body, violin and bow, permitted a kinematic analysis of position
changes resulting from the use of different shoulder rests. The study identified that, since the
instrument has a fixed geometry, the performer adapts to the instrument throughout the
learning process. Shoulder rests are one of the few tools available to adapt the instrument to
the performer. Rabuffetti et al concluded that ―optimization of the violin setup cannot be
based on objective evidence only (anthropometric size, violin features, type of repertoire, etc.)
but still requires a trial-and-error approach.‖ Motion analysis was used as means to provide
information that might accelerate this process (Rabuffetti, Converti, Boccardi, and Ferrarin,
2007).
In a study of trumpet mouthpieces, Devroop and Chesky (2002) examined
intraoralpressures, and compression of the mouthpiece against the embouchure. The basic
assumption was that these physical elements contribute to performance related medical
problems among trumpeters. Results reinforced concerns regarding risk for performance-
related medical problems, including rupture of the orbicularis oris muscle, tooth
displacement, head and neck lesions, embouchure dystonia, and temporomandibular joint
(TMJ) and jaw pain. In spite of the fact that many pedagogues and performers had attempted
to explain some of the mysteries surrounding mouthpiece forces the study contended that
their relationship was still not fully understood because these explanations were based upon
empirical evidence and not quantitative measurements. The study confirmed that laboratory-
based research can be effective in searching for answers to real-world problems of performing
artists (Devroop and Chesky, 2002).
In a 3-dimensional biomechanics study of pianists hand and finger movement patterns,
Virgilio et al. (2007) analyzed the movements recorded using a multi-camera motion capture
system. A complex musical sequence, 16 measures of a Bach Minuet, was used as a test
protocol for estimating the actual load imposed on joints during playing, as well as to assess
the influence of experience on the actual movements. The central hypotheses blended
concepts of injury prevention and pedagogy – that different patterns of movement would be
used depending on the player‘s experience and that pianists could use different techniques to
reach the same results. The study identifies that: 1) the movements required for piano playing
usually involve low impact loads that do not exceed physiologic limits of human body, but
their repetition may provoke microtrauma leading to overuse injuries; 2) Experience may
allow a pianist to modify the motor patterns used for a performance, allowing the highest
accuracy with minimum effort; and 3) that ―didactic techniques should always teach the
students how to reduce—or better, to avoid—professional disorders‖ and ―the use of a wrong
technique may accelerate the development of overuse stress injuries‖. Such work underpins
the premise that prevention is a matter of risk management during the teaching and learning
process (Virgilio, Chiara, Emilia, Paolo, and Sforza, 2007).
Together, these and the other available few examples all directly or indirectly suggest a
need for more research in the area of optimizing practice time and pedagogy with regard to
both efficiency (skill acquisition time) and effectiveness (learning while managing injury
risk).
3.5. Future Directions
Based on the current state of Arts Biomechanics research, there remains a considerable
amount of work to do. Bridging the gap between quantitative research and empirical
experience – thus providing vocationally meaningful research results – requires a marriage
between the quantitative powers of science and the empirical/vocational strengths of the
performing Arts. Multidisciplinary collaborations hold the potential to do exactly that.
Integrative multidisciplinary research can provide outcomes that have practical application by
co-opting the involvement and efforts of performers, teachers, medical practitioners, and
biomechanics researchers. Even though term ―multidisciplinary‖ has become catch-phrase of
some currency, multidisciplinary research is considerably more difficult to do than might
initially be expected. Perhaps this accounts for the dearth of research in the area. In Arts
Biomechanics, multidisciplinary research requires the integration of the sciences, fine arts and
education. Researchers must engage in dialogue about disciplinary perspectives throughout
the process, working directly with practitioners and accounting for perceptual experience at
all stages of the work. In short, it is a challenging but rewarding endeavour.
Arts Biomechanics has a particularly important role to play in the area of injury
prevention, risk management and remediation. The authors see this as a three stage process.
Context is established through integrating the multiple perspectives provided by MACRO
studies (kinematic, kinetic, and EMG research). Such methods reveal quantitative ―external‖
(motor control pattern and intensity of load) parameters and qualitative internal ones
(normalized EMG). They provide the first stage of a holistic approach to establish the true
nature of biological loading. Additional explorations of OS must include the examination of
context-relevant accumulations of micro-trauma on the MICRO level. This second stage of
biomechanics research involves applying MACRO conditions in MICRO-level research. We
speculate that existing technology can be used to determine the process of muscle fibre
damage and patterns of activity-specific muscular trauma. The third stage must examine
muscular regeneration under different circumstances. Recovery during stages of relative
inactivity (rest) is a vital factor for risk management.
Stage 1. Determination of Realistic Loading Conditions

3-D motion capture, biomechanical modeling, and EMG provide multiple viewpoints for
researchers and medical practitioners. Multi-dimensional quantitative analysis integrates their
several perspectives and allows a systematic evaluation of biological loading. The main
difficulties in this stage come from the high number and heterogeneity of signal components
in terms of sampling frequencies, units of measure, diversity of generally accepted practice in
dealing with the varied raw data, and interpretation of the results. Individually and
combinatorially, using several measurement techniques permits quantitative description of
factors related to the mechanism of injury. Kinematic modeling confirms group effects among
muscles through positional description (an external perspective) and muscle length changes
(an internal view). Dynamic and inverse dynamic modeling based on kinematic
characteristics establish quantity of load, and quantify some factors related to quality (impact,
fundamental, static, dynamic). EMG determines the presence and general level of loading
(e.g. relative muscle activity) with the normalization process pointing toward the
physiological limits of the individual. Muscle lengths in combination with EMG and kinetic
results create a clearer picture of the working pattern of the muscles. The main task in this
phase is to measure diverse Arts performance contexts.
Stage 2 – Micro Examination of Muscle Fibre Under Realistic Conditions from Stage 1
Imaging techniques such as LM, TEM, and MRI are frequently used to examine already
injured muscle areas but not the actual process of injury10 (Brandser, Elkhoury, Kathol,
Callaghan, and Tearse, 1995; Komulainen, Takala, Kuipers, and Hesselink, 1998; Rubin et
al., 1995; Uchiyama, Tamaki, and Fukuda, 2001). Traditional microscopic techniques have
long been used to categorize morphological changes in stressed muscle tissue. Until two
decades ago, the modalities available to image injuries for muscle and muscular attachments
were plain radiography and to a lesser extent ultrasound. Plain radiography is ineffective
because of its limited tissue contrast (Sundar and Carty, 1994). Ultrasound is highly operator
dependent and has a limited view of soft tissue. More recent developments include computed
tomography that can only obtain images in the axial plane and magnetic resonance imaging
(MRI). In particular, MRI has recently become the modality of choice for imaging muscles,
tendons and ligaments.
Muscle research using animal models shows that injuries can be detected using
histological and MRI methods (Biral et al., 2000; Kanbe et al., 1998; Smith, Plyley, Rodgers,
and McKee, 1997; Thompson, Balog, Fitts, and Riley, 1999). These studies focus on impact
(or near-impact) loading, static loads at high levels of intensity, eccentric (lengthening)
conditions, and end-state examinations of the muscles (Koskinen et al., 2002; Stauber and
Willems, 2002; Vijayan, Thompson, Norenberg, Fitts, and Riley, 2001). Such methods do not
mimic the conditions that lead to OS, where loads are normally far below ―impact‖ levels,
load patterns vary over time, and injuries are the result of micro-trauma accumulation. We
suggest that, using realistic loading conditions determined in MACRO research, MRI
techniques could be used to track the accumulation of micro-trauma using in-vivo methods,
examining muscle deterioration over longer loading durations.
Stage 3. Determination of Accumulated Micro-Trauma

Based on the current success of MRI in MICRO-level imaging of existing injuries (Rubin
et al., 1995; Steinbach, Fleckenstein, and Mink, 1994), a time course series could be used to
determine the effect of repetitive stimulation on a muscle, and the rate at which micro-trauma
accumulates. Micro-level studies may also be able to explore rates of repair during cycles of
rest. In Section 3.3 (The Theory of Biomechanical Modeling as It Applies to OS) OS was
mathematically represented time-dependent terms as follows:
CD = (I)ndt  MRdt
Since both of these factors involve internal biological processes, research must focus on
determining activity-specific values for I and n in the first of these integrals and on MR for
the second.
Prevention first requires meaningful measures and then comparison to the real world.
Arts Biomechanics permits the integration of information from multiple inputs with a
quantitative, time-based evaluation of load. It provides linking perspectives between the
process and the endstate of OS and is the only method currently available that can do so. It
remains to fully apply Arts Biomechanics in the service of determining factors associated
with OS.
4. APPLICATIONS OF MOVEMENT SCIENCE TECHNOLOGIES IN THE

CREATION OF ART
Movement science as a field can sometimes contribute to the arts in unexpected ways.
Because innovation is a driving force in both science and art, technology can create
conceptual bridges between Art and Science. In the previous sections of this chapter, Arts
Biomechanics was seen to provide a means to characterize intersections between body
movement and Artistic creation. Technology provided analytic tools and biomechanics
provided a means to understand and interpret the data. This relationship can be extended to
the creation of art, whether it involves actual movement or provides an analytic perspective of
existing artworks. Technology frequently carries potential beyond its intended applications,
but artists, composers, performers, and scientists need to work together in order to harness its
latent possibilities in the service of the creation of art. Ultimately audiences will be the judge
of an artwork‘s success or failure. Technology‘s role needs to be carefully considered so that
in does not merely serve as a layer of complexity, confounding communication between artist
and audience. Rather, the aim of technology should be to enhance the intensity and relevance
of the artistic concept. The purpose of this section is to provide discussion of several such
links.
In the visual arts, creative motivation typically derives from artists‘ emphasis of their
own individuality and their subjective interpretations of phenomena. As such, the end product
can be understood as an effort to represent different aspects of the human experience, as they
perceive it. This, in fact, is not just an artistic concept, it is a scientific one, too. Both science
and art are contingent upon having a means of ordering one‘s perception of the world and of
expressing that perception. Science does so with numbers and formulae while the visual arts
does so with images and sculpture. Using the approaches of science, movement analysis
technology can provide a link between artistic representations of human movement and our
experiential understanding of the bodily condition. The following discussion tracing some of
the influence of movement science on the visual arts is based on an in-press manuscript by the
authors (Shan, Visentin and Harnett, 2009).
Historically, representing time-based event of human movement in an image or sculpture
has been a challenge to artists. Pre-photography, human movement was typically represented
by capturing a single critical point during the unfolding of a gesture. Such stop-action
representation often exaggerated the phenomenon in order to emphasize an ―affect‖ that
would elicit empathy in the viewer; such art typically relies on the viewer‘s experiential
knowledge (of their own bodily condition) or on iconography and/or cultural loading.
Figure 20. Diskobolos of Myron (circa. 460 BC).
For example, Myron‘s classical representation of a discus thrower (circa. 460 BC) is,
from an anthropokinetical standpoint, hardly realistic. Yet, its representation of athleticism
and dynamism are understood; the contorted, unnatural body position as a means to show
power and effort has become iconographic (Figure 20).
"What work is there which is as distorted and elaborate as that Diskobolos of Myron? But
if anyone should criticize this work because it was not sufficiently upright, would he not
reveal a lack of understanding of the art, in which the most praiseworthy quality is this very
novelty and difficulty?" Lucian III: The Lover of Lies (1916) translated by A. M. Harmon
(Webpage of the University of Chicago, 2006).
Until the 20th century, affective representations of human movement are the norm.
Developments in technology – the advent of photography – inspire some of the most
profound changes in 2-D image-making. Photographic techniques were employed by both
scientists and artists. The work of Edward James Muybridge may have been scientific
investigations, but they had a huge impact on the art world. Muybridge used multi-camera
stop action photography (Figure 21 and 22) and multiple exposure techniques (Figure 23) to
initiate a scientific approach to human movement studies (Cavanagh, 1990; Solnit, 2003).
Further developments using photography to capture and analyze human movement can be
found in the work of: 1) Etienne-Jules Marey, who worked with high-speed photography at
the same time as Muybridge, capturing multiple images with one camera on one photo plate
(Braun, 1992); 2) Wilhelm Braune and Otto Fischer, who created elaborate 3-D models and
performed thorough gait analyses using 4 cameras and subjects marked with Geissler tubes
which emitted current that could then be seen on film (Cavanagh, 1990); 3) Nikolaj
Bernstein, who developed precise kinematic procedures using cyclography, a technique

employing film that was slowly moving through a camera with a mirror placed at a 45 degree
angle to the optical axis, thereby creating a picture that gave a frontal and lateral view of the
subject (Medved, 2001); and others – using techniques such as strobes, multiple and time
exposure (Plagenhoef, 1971). Biomechanical analysis using modern-day, motion capture
technology and can be considered a direct outgrowth of this work.
Directly influenced by photography, early 20th-century Futurist painters such as Umberto
Boccioni and Gino Severini began to experiment with these ideas, creating works (Figure 24,
Dynamism of a Dancer [1912]) that compressed the principle of multiple exposure imagery
into ―kaleidoscope art of force lines, vectored shapes and mobile patterns‖ (Hunter, Jacobus,
and Wheeler, 2004).
Figure 21. Multi-camera stop action recording of a pole-vaulter, by Muybridge.
Figure 22. Multi-camera stop action of a nude descending a staircase, by Muybridge.

Figure 23. Multiple exposure photography documenting human movement, by Muybridge.
Figure 24. Gino Severini - Painting: Dynamism of a Dancer [1912].

Displaying multiple fragmented essences in a single visual image, these artworks

approach an expression of time and position designed to represent a perception of the human
experience, by combining the traditional idea of stop-action affective representation with
Muybridge‘s multiple exposure techniques.
In each of these instances, the artwork can be understood to capture a sequential
expression of how we experience existence; one where technology has influenced our
existential perception of the world. American Modernist art sometimes strips down the
imagery to its bare essentials, leaving only gesture. For example, Franz Kline (Kline, 1950)
condenses gesture into a single frame of reference (one that may not be anthropokinetically
accurate, but is bodily expressive), and contemporary artist Tanya Harnett builds the
dynamics of a 3-D space through layered subject matter, gesture, and historical reference,
using them as a stage for a photographic image that compresses time, x-y-z dimensionality
and cultural reference into a flat-surface gestural work. Harnett‘s Nude Descending a
Staircase (Figures 25 and 26) is a direct reference to Marcel Duchamp‘s work of the same
title, who acknowledged the influence of Etienne-Jules Marey‘s motion capture studies.
When such works are viewed, they can be understood as a metaphors and commentaries on
the human experience.
Figure 25. Nude Descending a Staircase (Tanya Harnett, 2008, reprinted with permission of the artist).
4.1. 3D Motion Capture Technology on Visual Art
If photographic techniques tested the limits of visual culture in the early 20th century,
film, video and more recent developments in computers (3-D motion capture, virtual reality,
etc.) do so in ours. Cognition in these dynamic environments also relies upon the empathy of
the viewers.
Figure 26. Nude Descending a Staircase (Tanya Harnett, 2008, reprinted with permission of the artist).
These newer technologies are relatively easy to use and they diminish demands formerly
made on content creators; simply, because video media unfolds sequentially, it is easy for
viewers to believe in its verisimilitude. With respect to motion capture technology as an
extension of video media, four fundamental and consequential questions arise: 1) Can 3-D
technologies facilitate analysis and categorization of dynamic, time-based motions? 2) How
much representative information is needed in order to evoke an event‘s unfolding over time
(affects of human movement)? 3) What implications does the technology have for artistic
image making? And, 4) can we begin to speculate about cultural context as a factor
influencing gesture?
4.2. 3D Motion Capture and the Visual Arts: Theoretical Considerations
Gesture communicates information, the meaning of which is driven by empirical

experience. Perceptions of gesture as an unfolding of time-based events are intrinsically
integrated with the form and function of human movement. Since viewers of a 2-D image
have experienced their own bodily gestures as a time-based event, they are inclined to
empathize with any such implications in an artwork. The aesthetic experience and
understanding of a visual image is both spatial and temporal. Hence, visual communication is
linked to the personal experience of the viewer, where culture, iconography and/or visual
analogy can play a role.
Motion capture provides the means to accurately record movement in 3-dimensional
computer space. The technique differs significantly from actual photography. First, the
information gathered is not an actual image, rather it is a collection of trajectories generated
by pre-defined, user-determined points. Second, the viewer‘s choice of perspective is not
controlled by camera placement because it can be altered after the motion has been captured.
Images synthesized from motion capture trajectories may be used to analyze characteristics
that are objectively representational of time and bodiliness. These may be considered
signature elements of human movement.
4.3. Exemplary Experiments
Notwithstanding type of activity, trajectory-based images generated using motion capture

suggested qualities that may be an aid to establishing signature elements of human movement.
The uniqueness of each activity can be understood by relational emphasis on factors such as
stability (balance, body position/posture), athleticism (weight, speed, power/vigor), quality of
movement (control, smoothness/abruptness), and dimension of the movement (excursion).
Selected examples, described below, showed how 3-D motion capture could aid discourse
about perceptual characteristics of human movement. Examples are drawn from sports and
dance.
Figures 27 and 28 shows trajectories associated with a side kicking technique in Tae
Kwon Do. Its main characteristics were:
Stability: trunk trajectories had relatively small, near-linear excursions and they tended to be
clumped in one area of the image. The largest trajectories (legs and feet) visibly occurred
on a plane rather than in a volume. Clear presence of a centre, circled by rotational trunk
trajectories suggested weight, balance and a centripetal centre.
Athleticism: Long trajectory lines of the legs, particularly the feet contrasted the ―low-
energy‖, shorter lines of the trunk, lending the impression of a speed differential and a
directed power. Orientation of the foot and leg trajectories was in contrast to those of the
rest of the image, showing dynamism.
Quality of movement: Continuities of trajectories were smooth, showing controlled gestures,
except for acute-angle structures that occurred at the apex of the foot excursion. Abrupt
changes of the direction indicated thresholds of the activity or bodily limits where energy
of the movement was released. Generally compact excursions of trajectory lines showed
the movement to be occurring in a limited space, suggesting bodily control.
Figure 29 shows a soccer kick, viewed from above. Its key characteristics were:
Stability: Most trajectories appeared to be aligned in one general direction and have nearly
similar lengths. Relatively parallel to one another, trajectories (particularly trunk and
head) showed a state of controlled balance.
Athleticism and dimension: Swirling trajectories of the arms were related to energy and
athleticism. The double structure of these swirls indicated building and release phases of
the movement. It also gave scale to the overall gesture.
Quality of movement: Abrupt changes in trajectory direction (points of curvilinear
discontinuity) showed the release of pent-up power, a quality related to dynamism.
Figure 30 shows a signature generated by a Grand Jeté from classical ballet.
Stability and quality: A dense mass of trajectories, and a planar orientation in the central
trajectories of the image (trunk) were evidence of balance and control. These contrasted
trajectories (e.g. arm) where excursions had a rotational focal point (i.e. shoulder) and the
regularity of the curvilinear path clearly indicated fluidity and speed. These signature
characteristics can be related to balancing.
Figure 27. Side view of Tae Kwon Do side kick with body trajectory lines included (above). Sequential
stop-frame reconstruction of the movement (below).
Figure 28. Tae Kwon Do side kick (front-back view) with body trajectory lines.
Figure 29. Soccer maximal instep kick (top view) with body trajectory lines.
Figure 30. Ballet Grand Jété with body trajectory lines.
Figure 31 shows a fundamental skill of classical Indian dance. In Indian Dance, hand
gestures and steps are executed according to highly codified traditions.
Figure 31. Indian classical dance (front View) with body trajectory lines. Sequential stop-frame
reconstruction of the movement (below).
Different combinations of gestures and movements provide a huge vocabulary through

which stories can be depicted in surprising detail, perhaps intended to surpass the spoken
word (Barboza, 2006).
Stability and control: The clearly formal structure of trajectory lines showed controlled
movement. Layered densities of line showed elements of architectural symmetry,
balance, and regularity and consistency of movement. Centre of gravity was virtually
fixed in a horizontal plane. Such characteristics, particularly the symmetrical,
architectural references which were the result of highly codified training, could perhaps
be seen to be cultural indicators.
Athleticism and Quality: Continuous gesture trajectories along with no obvious points of
curvilinear irregularity indicated dynamism to be less important. Energy expenditure was
directed toward control rather than athleticism, which would have more indications of
gathering and release. Generally compact excursions of trajectory lines suggested
movements in a limited space.
For human movement, the experiential nature of bodiliness is a link connecting the
spatial and temporal. Activities such as sports, dance and performance are loaded with
cultural mythos, whether it is the embodiment of the warrior class, religious symbolism or the
inertia of cultural fashion. Movement signatures necessarily play a role in storytelling. This
assertion acknowledges important differences between the concepts of ―history‖ and
―aetiology‖. Whereas history is often viewed as developmental, which should not be
conflated with eschatological, aetiology investigates the conditions needed for historical
change to take place. Change occurs within the context of culture – whether biologically or
philosophically motivated.
Examining sports, the heritages harkens to a surrogate for warrior-class activities,
emphasizing power and speed, with additional elements such as control and balance. These
can be understood as signature elements of activities derived from a tradition of battle.
Culturally, sports teams are an extension of the concept of championing, where selected
representatives of the group pit their skills against opponents in proxy for the group as a
whole. The signatures of Figures 27 to 29 reveal these very characteristics.
Performing arts activities such as dance represent traditions that can also be understood
as a proxy; one for the values of the group heritage. In the case of activity 4, Grand Jeté,
cultural references are rooted in the Court of Louis XIV, where the foundations of classical
ballet originate and dance was used as a political tool. As a matter of note, when he took
power after reaching the age of majority, Louis the XIV, a virtuoso dancer, performed in an
allegorical ballet, dressed as the Sun God Apollo, symbolizing the rising of the sun to drive
away of the darkness of the previous political regime. From this point forward, Louis XIV‘s
noble class involved much of their efforts in learning ballet in order to imitate the noble
bearing of the King (Hilton, 1997). Hence, fundamental signature elements - upright bearing,
control, naturalness and demonstrated strength through elegant fluidity - become highly
codified; controlled symmetrical dance structures and steps analogous with the nobility of the
King and his natural, god-given place in the universe. These elements are still present in the
fundamentals of ballet today (Figure 30).
Clearly, many activities can be viewed as culturally loaded. The East Indian dance
signature (Figure 31) is visually distinctive, with characteristics that display highly codified
dance traditions – a stylization that captures something that is meaningful to or representative

of the culture. In East Indian dance, iconography is very direct. Much is related to storytelling
(religious or historical) through gesture. Formal, architectural structures, stability and control
can be related to the rhetorical cohesion needed to tell these stories. Sociological and/or
anthropological understanding enhanced by movement analysis technology might be one
bridge in the exchange of cross-cultural ideas. Understanding our experience as a temporal
bodily phenomenon creates cognitive links between information, still imagery, and cultural
mythos as other kinds of metaphors for the human experience.
4.4. Other Contemporary Examples
The potential of movement analysis technology in the creation of art has been explored in
several settings. Notably, Paradiso (1999) explored unconventional interfaces for individuals
to interact with musical environments. In his ―Brain Opera‖ project responsive environments
were built out of ―smart objects and sensitive spaces, where any kind of physical activity or
motion can result in a sophisticated multimedia reaction‖ (Paradiso, 1999). The kinds of
interfaces explored permitted interactive musical mappings based on the intuitive responses
of the participants (who were also the audience). Although the authors claim that several of
the Brain Opera's goals were achieved, the experiment has essentially proven to be a ―one-
off‖. It may have been an ambitious way of ―hinting at the possibilities now opening for
large-scale participatory music installations,‖ however the aim seems to have been mostly an
exploration of technological capacity and the project fell short of enhancing the intensity and
relevance of an artistic concept. In the study, Paradiso explained this to arise from the way in
which the ―project refuse(d) to fit into any clearly defined musical category‖. In a separate
study, Paradiso et al. (2003) further examined the concept of interactive musical
environments using electromagnetic tagging technology, where an ensemble of passively
tagged objects was tracked in real time. These were tied to simple notes and musical effects
which enabled dynamic control of musical sequences and sonic textures (Paradiso, Pardue,
Hsiao, and Benbasat, 2003). Although technologies can advance installation-based interactive
music environments, their communicative efficacy still rests in the ears of the audience.
In a multimedia project incorporating music and dance. Bevilaqua et al. (2001) used a 3D
motion capture system to produce animations from dance and, more unusually, generate a
soundtrack from the dancer's movements. Motion capture data was used to both trigger and
modify the timbre of sounds. The project was exploratory, challenging traditional boundaries
between the arts (music, dance, studio art and drama) and sciences (motion analysis
technology). The authors define their thesis as follows:
―In electronic music, the relationship between sound generation and touch/gesture is
artificially defined, contrary to the way in which sound is produced in acoustic instruments. In
fact, the possibility of designing the interface between the gesture (or the touch) and the sound
is a fascinating feature of digital music. Most electronic music controllers that have been
created are based on existing acoustic instruments (for example the piano keyboard). Such
electronic controllers have the obvious advantage to be used relatively easily by ‗traditionally‘
trained musicians‖ (Bevilacqua, Naugle, and Valverde, 2001).
In summary, such research consistently points to the need and desire for greater
collaboration between artists and scientists. In the given examples, technology provides
analytic tools and a means to characterize intersections between the body and artistic creation.
Biomechanical signals link the artistic conceptualization and its interface with our
perceptions. Such explorations are just starting. And, while the benefit may currently seem a
bit one-sided – artists taking advantage of technology for the betterment of their art – the
benefit really flows both ways. Technology is in a consistent state of rapid change. Artists‘
demands can be a catalyst for improving it. For example, in the case of 3-D motion capture,
some of the first demands for video and sound to be synchronized with the 3-D data came
from Arts researchers (VICON motion capture). This development is ubiquitous in the
industry today.
In terms of biomechanics research, there continues to be significant unexplored potential.
Throughout history, each set of technologies has ushered its own set of changes into the way
people interact with their environment and generate understanding. Technology may allow
complex mapping functions to be interpreted very sophisticated ways. But, it is sometimes
more interesting and useful to consider the artistic gesture, the potential for expression, or the
aesthetic design inherent in the digital data generated by it. This kind of thought is a central
tenant of the artistic process. And, to reconsider the first question posed in this chapter, it is
one of the best reasons to ―bother with Arts Biomechanics‖.
ACKNOWLEDGMENTS
This book chapter was financially supported by NSERC (National Sciences and
Engineering Research Council of Canada), Westgrid/Canada.
REFERENCES
Administration, F. A. (1996). Dynamic Evaluation of Seat Restraint Systems andOccupant
Protection on Transport Airplanes. Retrieved. from.
Bach, C. P. E. (1759). Essay on the True Art of Playing Keyboard Instruments (W. J.
Mitchell, Trans.). New York: W. W. Norton and Co.
Baillot, P. (1835). L'art du violon (L. Goldberg, Trans.). Paris: Northwestern University
Press.
Ballreich, R., and Baumann, W. (1996). Grundlagen der Biomechanik des Sports (The basics
of biomechanics in sports). Stuttgart: Enke Verlag.
Ballreich, R., and Kuhlow, A. (1980). Beiträge zur Biomechanik des Sports. Schorndorf: Karl
Hofmann.
Barboza, K. (2006). Repräsentative und symbolische Zeichen des klassischen Tanzes in
Indien und in Europa: Eine interkulturelle, vergleichende Analyse des indischen
bharatanāyam und des französischen Balletts. University of Münster, Münster.
Barth, T. (2000). An AIR passenger airbag system for transport aircraft. Retrieved. from.
Bevilacqua, F., Naugle, L., and Valverde, I. (2001). Virtual dance and music environment
using motion capture. Paper presented at the IEEE Multimedia Technology and
Applications Conference, Irvine.
Biral, D., Jakubiec-Puka, A., Ciechomska, I., Sandri, M., Rossini, K., Carraro, U., et al.
(2000). Loss of dystrophin and some dystrophin-associated proteins with concomitant
signs of apoptosis in rat leg muscle overworked in extension. Acta Neuropathologica,
100(6), 618-626.
Brandser, E. A., Elkhoury, G. Y., Kathol, M. H., Callaghan, J. J., and Tearse, D. S. (1995).
HAMSTRING INJURIES - RADIOGRAPHIC, CONVENTIONAL TOMOGRAPHIC,
CT, AND MR-IMAGING CHARACTERISTICS. Radiology, 197(1), 257-262.
Braun, M. (1992). Picturing Time: The Work of Etienne-Jules Marey (1830-1904). Chicago:
University of Chicago Press.
Brown, A. N. (1997). Musculoskeletal misuse among youth symphony string players.
Medical Problems of Performing Artists, 12, 15-18.
Cavanagh, P. (1990). The mechanics of distance running: a historical perspective. In P.
Cavanagh (Ed.), Biomechanics of Distance Running (pp. 135-164). Champaign: Human
Kinetics.
Chesky, K., Kondraske, G., Henoch, M., Hipple, J., and Rubin, B. (2002). Musicians‘ health.
In R. Colwell and C. Richardson (Eds.), The new handbook of research on music
teaching and learning. New York: Oxford University Press.
Chockley, C. (2008). Ground Reaction Force Comparison Between Jumps Landing on the
Full Foot and Jumps Landing en Pointe in Ballet Dancers. Journal of Dance Medicine
and Science, 12(1).
Dawson, D. M., Hallett, M., and Millender, L. H. (1990). Entrapment Neuropathies. Boston:
Little, Brown and Co.
Dawson, W. J. (2003). The bibliography of performing arts medicine: A five-year
retrospective review. Medical Problems of Performing Artists, 18(1), 27-32.
Dawson, W. J. (2007). The bibliography of performing arts medicine - A 10 year
retrospective review (Part II). Medical Problems of Performing Artists, 22(4), 153-159.
Devroop, K., and Chesky, K. (2002). Comparison of biomechanical forces generated during
trumpet performance in contrasting settings. Medical Problems of Performing Artists,
17(4), 149-154.
Einstein, A. (1931). The World as I See It. Retrieved January 9, 2009, from
http://www.aip.org/history/einstein/essay.htm.
Evans, B. (2003). How I Survived My Dance Training. Medical Problems of Performing
Artists, 18(4), 137-140.
Fotheringham W. (2002). Put Me Back on My Bike: In Search of Tom Simpson. New York:
Yellow Jersey Press.
Fry, H. J. H. (1986). Overuse syndrome of the upper limb in musicians. Medical Journal
Austrilia, 144, 182-185.
Fry, H. J. H. (1987). Prevalence of overuse (injury) syndrome in Australian music schools.
British Journal of Internal Medicine, 44, 35-40.
Fry, H. J. H. (1988). Patterns of over-use seen in 658 affected instrumental musicians.
International Journal of Music Education, 11, 3-16.
Fry, H. J. H., Ross, P., and Rutherford, M. (1988). Music-related overuse in secondary
schools. Medical Problems of Performing Artists, 3, 133-134.
Gadd, C. (1966). Use of a weighted impulse criterion for estimating injury hazard. Paper
presented at the 10th Stapp Car Crash Conference, New York.
Gadd, C. M. (1966). Use of a weighted impulse criterion for estimating injury hazard. Paper
presented at the 10th Stapp Car Crash Conference, New York.
Hagglund, K. L. (1996). A comparison of the physical and mental practices of music students
from the New England Conservatory and Boston University Music School. Medical
Problems of Performing Artists, 11, 99-107.
Hartsell, H. D., and Tata, G. E. (1991). A retrospective of music-related musculoskeletal
problems occurring in undergraduate music students. Pysiotherapy Canada 43(1), 13-18.
Hennig, E. M., and Lafortune, M. A. (1991). Relationships between ground reaction force and
tibial bone acceleration parameters. International Journal of Sport Biomechanics, 7, 303-
309.
Hilton, W. (1997). Dance and Music of Court and Theatre: selected Writings of Wendy
Hilton. Stuyvesant NY: Pendragon Press.
Hodgson, P. (1958). Motion Study and Violin Bowing. Urbana: American String Teachers
Association.
Hunter, S., Jacobus, J., and Wheeler, D. (2004). Modern Art, Third Edition Revised and
Expanded. New York: Prentice-Hall.
Kanbe, K., Hasegawa, A., Takagishi, K., Shirakura, K., Nagase, M., Yanagawa, T., et al.
(1998). Analysis of muscle bioenergetic metabolism in rabbit leg lengthening. Clinical
Orthopaedics and Related Research(351), 214-221.
Kaplinski, V. (2003). Prevention of injuries among keyboard players. Paper presented at the
Twenty-First Annual Symposium on Medical Problems of Musicians and Dancers.
Kline, F. (1950). Chief. Retrieved March 22, 2008, 2008, from http://www.moma.org/
collection/browse_results.php?object_id=78319
Kolneder, W. (1993). The Amadeus Book of the Violin - Construction, History, and Music.
Oregon: Amadeus Press.
Komulainen, J., Takala, T. E. S., Kuipers, H., and Hesselink, M. K. C. (1998). The disruption
of myofibre structures in rat skeletal muscle after forced lengthening contractions.
Pflugers Archiv-European Journal of Physiology, 436(5), 735-741.
Kosiak, M. (1959). Etiology and Pathology of ischemic ulcers. Archives of Physical
Medicine, 40, 62-69.
Koskinen, S. O. A., Ahtikoski, A. M., Komulainen, J., Hesselink, M. K. C., Drost, M. R., and
Takala, T. E. S. (2002). Short-term effects of forced eccentric contractions on collagen
synthesis and degradation in rat skeletal muscle. Pflugers Archiv-European Journal of
Physiology, 444(1-2), 59-72.
Kuipers, H. (1998). Training and overtraining: An introduction. Medicine and Science in
Sports and Exercise, 30, 1137-1139.
Lehmann, A. C., and Davidson, J. W. (2002). Taking an acquired skills perspective. In R.
Colwell and C. Richardson (Eds.), The new handbook of research on music teaching and
learning. New York: Oxford University Press.
Levitin, D. (2006). This is Your Brain on Music. New York: Plume.
Lockwood, A. H. (1988). Medical problems in secondary school-aged musicians. Medical
Lundberg, A. (1996). On the use of bone and skin markers in kinematics research. Human
Movement Science, 15(3), 411-422.
Magill, R. A. (2001). Motor learning concepts and applications (6 ed.). Boston: McGraw-Hill.
McLean, A., and Anderson, R. (1997). Biomechanics of closed head injuries. In P. Reilly and
R. Bullock (Eds.), Head Injury (pp. 25-37). London: Chapman and Hall.
Medved, V. (2001). Measurement of Human Locomotion. Boca Raton: CRC Press.
Middlestadt, S. E., and Fischbein, M. (1989). The prevelance of severe musculoskeletal
problems among male and female symphony orchestra string players. Medical Problems
of Performing Artists, 4, 41-48.
Moyse, M. (1973). The Flute and its Problems; Tone Development Through Interpretation for
the Flute. Tokyo: Muramatsu Gakki Hanbai Co.,Ltd.
Mozart, L. (1756). A treatise on the Fundamental Principles of Violin Playing (E. Knocker,
Trans.). Oxford: Oxford University Press.
Nemoto, K., and Arino, H. (2007). Hand and upper extremity problems in wind instrument
players in military bands. Medical Problems of Performing Artists, 22(2), 67-69.
Nicol, K. (1996). Velocity vs. duration of competition - A tool for characterizing sport events,
internal load, and limiting factors. Paper presented at the 9th Biennial Conference,
Canadian Society for Biomechanics, Vancouver.
Nicol, K. (1998). Gesetzmäßigkeiten von Grenzleistungen. Leistungssport, 5 28-35.
Palac, J. A. (1992). Violin bowing technique: an analysis of contemporary pedagogical
literature according to principles of human movement. Medical Problems of Performing
Artists, 7(1), 30-34.
Paradiso, J. A. (1999). The brain opera technology: New instruments and gestural sensors for
musical interaction and performance. Journal of New Music Research, 28(2), 130-149.
Paradiso, J. A., Pardue, L. S., Hsiao, K. Y., and Benbasat, A. Y. (2003). Electromagnetic
tagging for electronic music interfaces. Journal of New Music Research, 32(4), 395-409.
Parlitz, D., Peschel, T., and Altenmuller, E. (1998). Assessment of dynamic finger forces in
pianists: Effects of training and expertise. Journal of Biomechanics, 31(11), 1063-1067.
Peterson, L., and Renström, P. (1987). Verletzungen im sport (Injuries in Sports). Köln:
Deutcher Ärte-Verlag.
Plagenhoef, S. (1971). Patterns of Human Movement: a cinematographic analysis. Englewood
Cliffs: Prentice-Hall.
Quantz, J. J. (1752). Essay of a Method for Playing the Transverse Flute (R. E. R., Trans.).
London: faber and faber.
Rabuffetti, M., Converti, R. M., Boccardi, S., and Ferrarin, M. (2007). Tuning of the violin-
performer interface: An experimental study about the effects of shoulder rest variations
on playing kinematics. Medical Problems of Performing Artists, 22(2), 58-66.
Rubin, S. J., Feldman, F., Staron, R. B., Zwass, A., Totterman, S., and Meyers, S. P. (1995).
MAGNETIC-RESONANCE-IMAGING OF MUSCLE INJURY. Clinical Imaging,
19(4), 263-269.
Schwartz, B. (1983). Great masters of the violin. New York: Simon and Schuster.
Shan, G. (2005). Comparison of repetitive movements between ballet dancers and martial
artists ¾ risk assessment of muscle overuse injuries and prevention strategies. Research
in Sports Medicine, 13(1), 63-76.
Shan, G. B. (1999). Ein biomechanisches Modell auf Basis von individuellen
anthropometrischen Daten fuer das Bewegungslernen der Flugphasen bei sportlichen
Bewegungsablaeufen (A biomechanical model for motor learning of air-borne phases
based on individual anthropometry). Müster: Lit Verlag.
Shan, G. B., and Bohn, C. (2003). Anthropometrical data and coefficients of regression
related to gender and race. Applied Ergonomics, 34(4), 327-337.
Shan, G. B., and Visentin, P. (2003). A quantitative three-dimensional analysis of arm
kinematics in violin performance. Medical Problems of Performing Artists, 18(1), 3-10.
Shan, G. B., Visentin, P., and Schultz, A. (2004). Multidimensional signal analysis as a
means of better understanding factors associated with repetitive use in violin
performance. Medical Problems of Performing Artists, 19(3), 129-139.
Shan, G. B., Visentin, P., Wooldridge, L., Wang, C. D., and Connolly, D. (2007). A
frequency-based characterization of spiccato bowing in violin performance. Perceptual
and Motor Skills, 105(3), 1027-1051.
Shippen, J., and May, B. (2008). Whole body biomechanical analysis and muscle loading of
Irish dancers. Paper presented at the International Association for Dance Medicine and
Science, 18th Annual Meeting, Cleveland, OH.
Shippen, J., and May, B. (2009). An analysis procedure for the calculation of muscle loading
and joint contact forces in dancers. Journal of Dance Medicine and Science.
Smith, H. K., Plyley, M. J., Rodgers, C. D., and McKee, N. H. (1997). Skeletal muscle
damage in the rat hindlimb following single or repeated daily bouts of downhill exercise.
International Journal of Sports Medicine, 18(2), 94-100.
Solnit, R. (2003). River of Shadows: Eadward Muybridge and the Technological Wild West.
New York: Viking Penguin.
Solomon, R., and Solomon, J. (2004). Publications in dance medicine and science: A
bibliographer's perspective. Medical Problems of Performing Artists, 19(4), 167-169.
Stauber, W. T., and Willems, M. E. T. (2002). Prevention of histopathologic changes from 30
repeated stretches of active rat skeletal muscles by long inter-stretch rest times. European
Journal of Applied Physiology, 88(1-2), 94-99.
Steinbach, L. S., Fleckenstein, J. L., and Mink, J. H. (1994). MAGNETIC-RESONANCE-
IMAGING OF MUSCLE INJURIES. Orthopedics, 17(11), 991-999.
Sundar, M., and Carty, H. (1994). AVULSION FRACTURES OF THE PELVIS IN
CHILDREN - A REPORT OF 32 FRACTURES AND THEIR OUTCOME. Skeletal
Radiology, 23(2), 85-90.
Taylor, J. (1996). Intensity and athletic performance. In J. Van Raalte and B. Brewer (Eds.),
A practitioner's guide to sport and exercise psychology. Washington: American
Psychological Association.
Thompson, J. L., Balog, E. M., Fitts, R. H., and Riley, D. A. (1999). Five myofibrillar lesion
types in eccentrically challenged, unloaded rat adductor longus muscle - A test model.
Anatomical Record, 254(1), 39-52.
Tromlitz, J. G. (1791). The Virtuoso Flute Player (A. Powell, Trans.): Cambridge.
Tulchinsky, E. (1995). A BIOMECHANICAL MOTION ANALYSIS OF THE VIOLINISTS
BOW ARM (VOL 9, PG 119, YR 1994). Medical Problems of Performing Artists, 10(1),
10-10.
Tulchinsky, E., and Riolo, L. (1994a). A biomechanical motion analysis of the violinist‘s bow
arm. Medical Problems of Performing Artists, 9(4), 119-124.
Tulchinsky, E., and Riolo, L. (1994b). A BIOMECHANICAL MOTION ANALYSIS OF
THE VIOLINISTS BOW ART. Medical Problems of Performing Artists, 9(4), 119-124.
Uchiyama, Y., Tamaki, T., and Fukuda, H. (2001). Relationship between functional deficit
and severity of experimental fast-strain injury of rat skeletal muscle. European Journal of
Applied Physiology, 85(1-2), 1-9.
USA. Airworthiness Standards: Transport Category Airplanes. Retrieved. from.
USGovernment. Airworthiness Standards: Transport Category Airplanes. Retrieved. from.
Vijayan, K., Thompson, J. L., Norenberg, K. M., Fitts, R. H., and Riley, D. A. (2001). Fiber-
type susceptibility to eccentric contraction-induced damage of hindlimb-unloaded rat AL
muscles. Journal of Applied Physiology, 90(3), 770-776.
Virgilio, F. F., Chiara, M., Emilia, B., Paolo, P., and Sforza, C. (2007). Three-Dimensional
Analysis of Hand and Finger Movements during Piano Playing. Medical Problems of
Performing Artists, 22(1), 18-23.
Visentin, P., Shan, G., and Wasiak, E. B. (2008). Informing music teaching and learning
using movement analysis technology. International Journal of Music Education, 26(1),
73-87.
Visentin, P., and Shan, G. B. (2003). The kinetic characteristics of the bow arm during violin
performance: An examination of internal loads as a function of tempo. Medical Problems
of Performing Artists, 18(3), 91-97.
Visentin, P., and Shan, G. B. (2004). An innovative approach to understand overuse injuries:
Biomechanical modeling as a platform to integrate information obtained from various
analytic tools. Medical Problems of Performing Artists, 19(2), 90-96.
Wilson, F. (1986). Tone deaf and all thumbs? New York: Viking Press.
Winter, D. A. (1990). Biomechanics and Motor Control of Human Movement. New York:
Wiley.
Zaza, C. (1992). Playing-related health problems at a Canadian music school. Medical
Zaza, C. (1998). Playing-related musculoskeletal disorders in musicians. Canadian Medical
Association Journal, 158, 1019-1025.
Chapter 2
UNDERSTANDING CORNEAL BIOMECHANICS

THROUGH EXPERIMENTAL ASSESSMENT AND
NUMERICAL SIMULATION
Ahmed Elsheikh
Ocular Biomechanics Group
Division of Civil Engineering, University of Dundee, UK
ABSTRACT
The Ocular Biomechanics Group was established in 2002 with one clear target; to
develop a virtual reality model of the human eye that can be used effectively and reliably
to predict ocular response to surgery, injury and disease. This ambitious, and seemingly
illusive, target helped plan our activities over the last 6 years and will still be focusing
our efforts as we strive to create the necessary knowledge using experimental methods,
build the predictive tools using programming and analysis means, and validate the
findings in both the laboratory and the clinic. This chapter presents an overview of our
biomechanical studies from laboratory material characterisation to finite element
numerical simulation. The chapter describes what has been achieved and points at the
remaining gaps in our knowledge. It explains that while much remains unknown in ocular
behaviour, we are now in a good position to use available knowledge to progress
predictive modelling and use it in actual applications such as improving the accuracy of
tonometry techniques, planning of refractive surgeries and design of contact lenses. The
discussion focuses on the cornea, although scleral biomechanics receive some mention.
The chapter also refers to microstructural, biomechanical and topographic studies
conducted by other research groups. Coverage of these studies has been necessary to
provide a more complete image of current understanding of corneal biomechanics.
INTRODUCTION
The transparent cornea is a most important component of the outer ocular tunic. It
provides a tough protective envelope for the ocular contents and helps give the eye its general
58 Ahmed Elsheikh
shape, Figure 1. The anterior corneal surface accounts for about two-thirds of the optical
power of the eye [1].
Cornea
Limbus
Lens
Sclera
Optic nerve
Figure 1. Cross-section of eye globe showing some of the main components [4].
This important role can be defined in terms of corneal shape, regularity and clarity, and is
a function of its refractive index. The cornea‘s contribution to ocular image formation can be
degraded by anomalies in shape produced by disease, injury or surgery. The ability to
understand and predict corneal response to such conditions is therefore of great clinical
importance [2, 3].
The past three decades saw a rapid growth in the use of numerical modelling to improve
understanding of corneal biomechanics and to predict ocular response to the conditions met in
tonometry and refractive surgery. Numerical models have improved in accuracy and in their
representation of the natural conditions of the eye. In the 1980s and 1990s, it was common to
assume linear elastic, and homogenous material behaviour and spherical topography in
corneal numerical models [5-8]. Since then, and with the better understanding of corneal
behaviour and topography achieved experimentally and clinically, these assumptions have
become unnecessary [9]. The increase in computing power over these three decades also
made it possible to produce and execute highly nonlinear corneal models with fine finite
element meshes and with consideration of other ocular components [10, 11]. These highly
expensive models, which required specialised mainframe computers in the past, can now be
conducted on commonly available personal computers.
However, in spite of this progress, corneal numerical simulation had only a limited role in
improving clinical practice. An example of this situation can be seen in the attempts to create
numerical models of refractive surgery, where much attention has been given to simulate the
procedures [12-23], but as yet no numerical model has been adopted in clinical practice to
plan and optimise refractive surgeries. Offering a solution here might not be straightforward
since part of the problem is related to the clinicians‘ reluctance to adopt tools and findings
that come through a numerical route. However, much progress can be made by reviewing
current modelling approaches, and concentrating on the areas where improvement is needed.
Understanding Corneal Biomechanics through Experimental Assessment… 59
Recent developments in corneal microstructure detailing, material characterisation and

topography mapping are offering valuable information for numerical modelling.
Microstructural studies have identified the preferential orientation of stromal collagen fibrils
[24-27] and the resulting anisotropy in corneal biomechanical behaviour [28, 29]. The corneal
nonlinear stress-strain behaviour [14, 30, 31], viscoelasticity [32, 33], hysteresis [34], long-
term creep [35] and stress relaxation [36] have been measured experimentally. The
epithelium‘s contribution to corneal biomechanics has been quantified in relation to the
stroma [37, 38]. More recently, the inter-lamellar adhesion within the stroma, under both
shear [39] and lateral tension [40-42], has been assessed. Much work has also been done to
determine the anterior and posterior asphericity [43-45], and the thickness distribution of the
cornea [46, 48]. These properties offer new opportunities to make numerical modelling more
representative of corneal behaviour, and potentially more reliable. But the question, which
this chapter attempts to answer, is how best to use this knowledge, and whether it is essential
for a numerical model to consider all these properties regardless of the application at hand.
The chapter starts with an overview of recent microstructural, topographical and
experimental advances that provided significant improvements in our understanding of
corneal biomechanics. A discussion is then presented on how to embed the experimentally-
obtained material behaviour in numerical simulation. Within this discussion, a general
introduction to finite element modelling and the simplifications involved in model
construction is presented, followed by an assessment of the effects of simplifications on
model output and whether these effects are acceptable. The chapter ends with concluding
statements on the way forward for corneal numerical simulation.
CORNEAL MICROSTRUCTURE
The human cornea consists of three distinct layers: an anterior epithelium, a posterior
endothelium and a central stroma, Figure 2. Between these layers are specialised extracellular
structures called Bowman's layer and Descemet's membrane. While the epithelium and
endothelium are cellular, composed mainly of keratinocytes, the stroma has a distinctive
composite structure with collagen fibrils embedded in an extracellular matrix rich in
proteoglycans, glycoproteins and keratocytes.
Anterior surface Epithelium
Bowman’s layer
Stroma
Posterior surface Descemet’s membrane
Endothelium
Figure 2. Cross-sectional view of the human cornea‘s layered construction.

60 Ahmed Elsheikh
Figure 3. Interweaving of stromal lamellae.
The mechanical behaviour of the human cornea is dominated by the stroma that forms
about 90% of its thickness [24, 49]. The stroma has a highly complex structure with 200-400
superimposed lamellae, set mostly parallel to the corneal surface, but increasingly interwoven
towards the anterior surface [25, 50-52], Figure 3. The increased lamella interweaving
possibly leads to variation in corneal mechanical properties and swelling resistance [53].
Stromal lamellae are composed of fine collagen fibrils regularly spaced and embedded in
an extracellular matrix. The fibrils within each lamella run parallel to each other, but make
large angles with neighbouring lamellae, Figure 4. Earlier research identified two preferential
directions within which most fibrils run in the central human cornea, namely the superior-
inferior and the temporal-nasal directions, possibly related to the action of the extraocular
muscles [26, 27, 54-57]. Daxer and Fratzl estimated that 49% of fibrils are oriented in these
two directions and 66% of the fibrils lie within the 45° sectors surrounding them [58]. Since
the fibrils are thought to be the main load carrying elements of the stroma, their preferential
orientation is likely to lead to anisotropic mechanical behaviour of the human cornea with the
vertical and horizontal directions exhibiting superior stiffness (resistance to deformation)
compared with other directions.
Figure 4. TEM images of (a) collagen fibril distribution within a stromal lamella, and (b) three lamellae
with different fibril orientations.
More recent studies found that corneas from other species exhibited significantly
different fibrillar collagen structures [59, 60]. The central corneas of horses, cows and
marmosets were found to have dominantly superior-inferior (vertical) orientation, while
porcine, rabbit and mouse corneas had mainly circumferential orientation. This variation in
structure was thought to be due to differences between species in the length, width and
insertion position of the rectus muscles, and possibly also due to the existence of the posterior
retractor bulbi muscle in all the above mentioned animals but not in humans [61]. However,
despite these differences, a common feature that exists in both human and animal corneas is
the gradual change in fibril orientation towards a predominantly circumferential organisation
close to the limbus.
Other corneal layers include the endothelium; a monolayer of specialised cells that line
the posterior surface of the cornea and face the eye‘s anterior chamber [62]. The endothelium
regulates fluid transport across the corneal posterior surface by allowing leakage of solutes
and nutrients from the aqueous humor to the cornea and pumping water from the stroma to
the aqueous [63]. The small thickness of the endothelium (≈ 2% of corneal thickness), and its
cellular structure suggest it has low mechanical stiffness (resistance to deformation)
compared to the stroma, although its stiffness has never been quantified [64].
Lining the other surface of the cornea is the epithelium which comprises about 8-10% of
corneal thickness [64]. The epithelium functions primarily to block the passage of foreign
material into the eye, absorb oxygen and nutrients from the tear film [1] and control the
swelling of the stroma [65]. Similar to the endothelium, the epithelium has a cellular
structure, which makes its mechanical stiffness low in comparison to that of the stroma.
Between the stroma and the endothelium lies Descemet's membrane, whose physiological
function is unknown although it may be to provide mechanical support to the endothelium, or
to act as a fluid barrier. Its low mechanical stiffness and small thickness (ranges between 3
and 10 μm or 0.5 to 2% of corneal thickness) suggest it has low contribution to the cornea‘s
overall mechanical behaviour [66, 67].
Finally, Bowman's layer is located between the epithelium and the stroma, composed of
strong collagen fibers and has a thickness of 8-12 μm. Similar to Descemet‘s membrane, the
physiological function of Bowman‘s layer is unknown.
It appears to provide strong attachment of the epithelium to the stroma in human eyes, but
the fact that it does not exist in some mammal eyes has not been explained [68]. Earlier
research assessed the mechanical behaviour of Bowman‘s layer and found it to be
insignificant [69].
The discussion in this chapter will assess evidence relating the corneal microstructure,
and in particular the stromal lamellae organisation, to overall biomechanical behaviour. The
implications of this relationship on our understanding of the effects of disease, injury and
surgery will also be covered in the conclusion section.
CORNEAL TOPOGRAPHY
The normal cornea is prolate – steepest at the centre and flattens towards the periphery. A
section through the cornea is an ellipse that can be described using Bennett‘s equation in
terms of two parameters; the radius of curvature at corneal apex (apical radius, Ro), and the
asphericity or shape factor, p:
62 Ahmed Elsheikh
R o  R o2  p x 2
z (1)
p
where x and z are coordinates of a general point, a, on the corneal surface, Figure 5. The
apical radius, Ro, (which approximately equals the value measured in keratometry) has been
measured in a few clinical studies, and examples of the reported average and standard
deviation values include 7.87±0.27 and 7.77±0.20 mm [43, 44]. On the other hand, the shape
factor, p, provides an indication of how rapidly the surface flattens away from the apex
(centre). The average values of p for the human cornea‘s anterior and posterior surfaces have
been found using in-vivo measurements to be 0.82 and 0.62, respectively [43].
Z
Co
rn
eal
su
r fa Ro a
ce
z
X
O x
Figure 5. Cross section through a corneal surface with coordinates of a general point a.
CORNEAL THICKNESS DISTRIBUTION

The cornea has a variable thickness that changes from minimum at or close to the centre,
to maximum along the limbus, Figure 6. Clinical studies reported different average values,
standard deviations and ranges of central corneal thickness, CCT, including 58054 µm
(range = 448-713 µm) [70], 548.229.6 (493-617 µm) [71], 53531 (490-567 µm) [72] and
551.34±40.58 µm for women and 542.34±43.84 µm for men [73].
Central corneal thickness (CCT)
Peripheral corneal
thickness (PCT)
Figure 6. Central and peripheral corneal thickness.

On the other hand, the average values of the peripheral corneal thickness, PCT, reported
in the literature varied according to where it had been measured. The values ranged between
600 and 700 µm for measurements taken between 15° to the corneal centre, and at the limbal
periphery [74-77]. The relationship between CCT and PCT is yet to be established, although
it is known that PCT is usually larger than CCT.
CORNEAL BIOMECHANICAL BEHAVIOUR

The composite microstructure of the cornea makes its biomechanical behaviour quite
complex and poses several difficulties including:
1. With the collagen fibrils being the main load carrying components of the stroma, the
behaviour becomes hyperelastic with an initially low stiffness (resistance to
deformation) increasing gradually under higher loads.
2. The preferential orientation of stromal lamellae leads to anisotropic behaviour with
the mechanical stiffness expected to be highest in the preferred directions of
lamellae.
3. The change of lamella orientation from the central area to become circumferential
close to the limbus affects the directional anisotropy of the tissue at the transition
zone.
4. The cellular composition of the epithelium and endothelium is expected to result in
significantly lower stiffness levels compared to the fibrous stroma.
5. The increased corneal thickness away from the centre suggests an associated increase
in both flexural and membrane stiffness.
6. Stromal swelling, common during sleeping and following prolonged wear of contact
lenses, increases the spacing between collagen fibrils and may affect the
biomechanical behaviour of the tissue especially under loads creating flexural
actions.
7. The increase in lamella interweaving in anterior stroma, and the reported effect of
interweaving in increasing swelling resistance, is expected to lead to variation in
biomechanical behaviour between anterior and posterior stroma [78].
8. Diseases such as keratoconus are known to affect the cross linking between collagen
fibrils in the stroma and lead to deterioration in corneal stiffness in the affected area.
The correlation between the reduced cross-linking in keratoconus and change in
mechanical properties requires further study [79].
9. The fast remodelling of the epithelium under external loads, such as those
experienced while wearing orthokeratology and other types of rigid contact lenses,
makes it difficult to study the cornea‘s long-term behaviour as in this case epithelial
remodelling and corneal creep take place simultaneously.
The discussion below presents the results of recent research studies conducted to
characterise corneal tissue and to address some of the challenges described above. In spite of
the significant advances that have already been made, current knowledge and understanding
of corneal biomechanical behaviour remain imperfect and further progress is needed.
64 Ahmed Elsheikh
EXPERIMENTAL ASSESSMENT OF CORNEAL BEHAVIOUR

Several experimental studies have been conducted since the 1960s to characterise the
biomechanical behaviour of corneal tissue [7, 36, 80-90]. The material constitutive models
(stress-strain and Young‘s modulus-strain relationships) reported in these studies have been
compared in Figure 7 and show a wide range of variation [91].
Figure 7. Biomechanical behaviour of corneal tissue as reported in earlier experimental studies, (a)
Stress-strain behaviour, (b) Relationships between tangent Young‘s modulus and strain – Studies
reported include Kampmeier [36], Nash [79], Hoeltzel [92], Bryant [7], Woo [87], Zeng [85] and
Woolensak [95].
This wide range, which is far beyond what could be acceptable in any predictive analysis
of corneal behaviour, is thought to be related to the viscoelasticity and anisotropy of corneal
tissue, and the effects of ageing and hydration. Studies adopting different strain rates or
employing different testing techniques reported different material properties [79, 85, 88, 92].
Specimens with varying donor ages were sometimes combined in the same test group despite
the known age-related changes in the microstructure of corneal tissue (including an increase
in stromal fibril diameter [93] and interfibrillar cross-linking [94]) and the associated effect
on the tissue‘s biomechanical behaviour. Corneal specimens with different swelling extents
were also used in earlier studies despite the current incomplete understanding of the effect of
hydration and swelling on corneal biomechanical behaviour.
For this reason, it has been essential that the characterisation of corneal biomechanical
behaviour be based on experimental techniques that offer tight control of strain rate, hydration
and temperature, consider the effect of age on tissue stiffness and test the cornea in a state that
closely simulates its normal physiologic conditions. Examples of these techniques are
described in the next section.
EXPERIMENTAL TECHNIQUES
Two main experimental techniques are currently available to assess corneal behaviour.
The first technique, called inflation testing, relies on supporting corneal buttons along the
limbus, applying uniform pressure on the cornea‘s posterior surface, and monitoring the
resulting deformation using non-contact methods, Figure 8. The pressure-deformation (p-)
data obtained experimentally are then converted into stress-strain (-) behaviour of corneal
tissue using shell mathematical analysis through the following steps (see Ref 96 for more
details).
6 5
8
1 = Corneal button
2 = Pressure chamber
3 = Water tank
4 = Temperature controller
4 9 5 = pressure transducer
6 = Loading reservoir
2 7 = Laser beam
1 7 8 = Camera
3 9 = Wetting device
Figure 8. Main components of an inflation test rig.
The secant Young‘s modulus, E, is first determined for a particular pressure-rise (p-)
behaviour point as:
1  ν  1  e-βη cos βη,

2
p Rm
E (2)
2δt
66 Ahmed Elsheikh
where Rm is the radius of the corneal median surface, t the average thickness assumed = ½
(CCT + PCT), η half the central angle of curvature (see Figure 9), η  sin
1
R i /R ant , Ri
radius of the corneo-scleral intersection, Rant radius of the cornea‘s anterior surface,
β  R/t  4 3 1  ν 2  , and  Poisson‘s ratio, taken as 0.49 assuming corneal tissue behaves
as an almost incompressible material [5, 7, 36]. The strain at this behaviour point is defined
as:
ε
p Rm
1  ν  1  ν e βη cosβ , (3)
2Et
and the stress as:
ζ  ε E. (4)
Anterior surface CCT = central

corneal thickness
Median surface
t = average thickness
= (CCT + PCT)/2
PCT = peripheral
corneal thickness
Ri
Posterior surface
Rm
Rant
Figure 9. Cross-sectional view of a corneal button showing the main dimensions used in the shell
mathematical analysis.
The inflation test maintains the cornea close to its normal working conditions and is
therefore considered suitable to determine the global stress-strain behaviour of corneal tissue.
The only deviation from normal in vivo conditions is in the development of a pinned edge
along the cornea‘s limbus, which in effect considers the sclera to be stiffer than the cornea to
the extent that it can be approximated as a rigid material. This approximation, which is
expected to have only a negligible effect on corneal behaviour, has been necessary to make
possible the mathematical shell analysis of corneal behaviour, which produced Equations 2-4.
A possible development of inflation testing involves the use of intact eye globes [81],
which could be supported along the equator (without preventing radial deformation) and
subjected to intraocular pressure loading through a canola inserted into the anterior chamber,
Figure 10. However, since no mathematical solution is available to convert the
experimentally-obtained pressure-deformation behaviour into stress-strain relationships, the

usefulness of this technique is still limited.
Figure 10. Loading and boundary conditions applied to (a) corneal button specimens, and (b) whole eye
globes subjected to inflation tests.
The second experimental technique is strip extensometry testing, in which a strip of

corneal tissue with a constant width is extracted and attached to the grips of a tension machine
while monitoring its behaviour. The stresses (ζ) and strains (ε) are calculated from the load
(T) and elongation (δ) data using the following simple equations:
ζ  T / A and ε  δ / L , (5)
where A and L are the specimen‘s original cross-sectional area and length, respectively. The
simplicity of the technique makes it quite common in ocular biomechanics research but a
number of inherent deficiencies, which can reduce its reliability, must be recognised and
accounted for. The deficiencies originate mainly from the initially curved form of the
specimen (Figure 11), the non-uniform corneal thickness and the termination of fibrils which
are not aligned with the specimen orientation.
The straightening of the specimen from its curved form results in initial strains that affect
the behaviour under subsequent loading [97], Figure 12. The specimen‘s curved form also
means that its edges are slightly shorter than the longitudinal centre line and this length
variation leads in turn to non-uniform distribution of stress under loading. The variable
corneal thickness, from a minimum at the centre to a maximum at the ends, further
complicates the derivation of the stress-strain relationship from the test load-elongation
results. Additionally, the relatively large thickness of the specimen poses another difficulty
with the potential of unequal clamping of external and internal corneal layers.
68 Ahmed Elsheikh
Cornea
Co
Corneo-scleral
rn
ea
intersection
l
tis
Scleral tissue
sue
ue
t i ss
al
le r
Sc
Figure 11. Initial curved form of corneal strip specimens.
Figure 12. Straightening of (a) an initially curved strip specimen produces stresses and strains with the
distribution shown in (b).
While these limitations affect the technique‘s suitability in obtaining the global material
properties of the cornea, the technique remains viable for comparative studies such as those
on anisotropy (variation in tissue behaviour between different anatomical directions) and
viscoelasticity (variation in behaviour with changes in strain rate). In studies on global
corneal behaviour, use should be made instead of the above described inflation testing.
However, mathematical techniques have been developed to consider some of the
deficiencies of strip testing [98], and these techniques could be useful where strip testing is to
be used to produce an approximate estimate of the stress-strain behaviour. The effects of the
specimen length variation and the straightening of the initially curved specimen form can be
handled first together. Figure 13 shows a 3D view of a strip extracted from a curved cornea
(assumed spherical for simplicity) and orthogonal cross-sections showing the variation of
specimen length from maximum along the centre line to minimum at the sides.
A
w L
Specimen centre line

L‘
H Specimen edge
R‘ Rm Rm R‘
Rm R‘
y 
O‘ O O‘‘
O, O‘, O‘‘
s s
A
Cross-section A-A Front Elevation
Specimen edge
Corneo-scleral
Specimen centre line intersection
Corneo-scleral intersection
Specimen edge
B
Specimen edge
Cornea
Specimen centreline
w z
A
Rm Specimen edge
R‘

Clamped end Monitored length Clamped end
O
 w
O‘
3D view Plan
Figure 13. Orthogonal and 3D views of a strip specimen extracted from a corneal button.
Reference is also made to Figure 14 which shows the strain distribution due to (a) the
variation in length, (b) initial straightening, and (c) the combined effect. Consideration of
these two factors starts with a modified form of the strain, ε (the derivation of Equations 6 to
9 can be found in Ref. 98):
δ h
ε  (6)
2 α  Rm  z 2 2 Rm
where δ is the measured specimen elongation under tension load T, 2α the angle of curvature
of the specimen, Figure 13, and Rm the radius of the corneal median surface. Distances h and
z are depicted in Figures 13 and 14b, respectively. This form of strain is then used in an
equation relating the applied force, T, and the resulting strain, ε:
T
 t/2  w/2

 t/2  w/2
a ε 3

 b ε 2  c ε  dz dh (7)
70 Ahmed Elsheikh
where a, b and c are constants to be determined such that the best fit with the experimental
data is achieved. With the values of the constants determined, the corresponding stress is then
calculated as:
ζ  a ε3  b ε 2  c ε (8)
t +h
+ =
n -h
t io io n on
i bu ut u ti
di st r tr ib t r ib
z d is d is
 d
d
z +
(a) Strain due to axial loading (b) Strain due to straightening of (c) Total strain
distribution is not uniform due to initially curved specimen
variation in specimen length
Figure 14. Strain distribution on the cross-section of a strip specimen.
The effect of thickness variation from minimum at the centre, t1, to maximum at the
clamps, t2, can be considered next in the stress calculations to develop a modified stress, ζ‘,
of the form, Figure 15:
t1   t2 L   t L 
ζ'  ζ  ln    ln  1  , (9)
t 2  t1    2   2 
where L is the specimen‘s initial length, w the specimen width, and ζ the stress derived in
Equation 8. An assumption of linear variation between t1 and t2 is adopted to simplify the
development of Equation 9. The second term of the equation:
t1   t2 L   t L 
ln    ln  1 
t 2  t 1    2   2 
represents the correction caused by the variable thickness. The second term is below 1.0 and
reaches this value only when t2 = t1.
With these formulations, some of the deficiencies of strip testing could be addressed in
the stress and strain calculations.
Other deficiencies such as the termination of some collagen fibrils along the specimen
sides and the difficulty in clamping the layered corneal tissue cannot currently be accounted
for using similar mathematical techniques.
Clamp CL Clamp
tlimbus tlimbus
t2 t1 t2
Clamp Clamp
L/2 L/2
Sclera Cornea Sclera
Figure 15. Thickness variation of corneal strip specimens.
INFLATION TEST SETUP

Inflation testing is based on supporting corneal buttons along their limbal tissue, applying
uniform posterior pressure that simulates the intraocular pressure and measuring the resulting
corneal deformation. The pressure-deformation experimental results are then converted
mathematically, using Equations 2-4, to stress-strain tissue behaviour.
An inflation test rig has been developed by the Ocular Biomechanics Group to enable
tight control of the conditions thought to have an influence on corneal behaviour including
hydration, temperature and load application rate. The rig is totally computer-controlled with
all test conditions specified electronically by the user and all behaviour monitoring devices
storing their data automatically for later analysis [30].
Fresh and well preserved corneal buttons are mounted onto the custom-built pressure
chamber shown in Figure 16 using only mechanical clamps. The pressure chamber is filled
either with saline solutions in tests lasting less than 45 minutes, or with the preservation
medium Optisol (which can maintain tissue hydration and prevent degradation for up to 14
days post mortem [99-102]) in longer tests. The chamber is connected to a reservoir, which
moves vertically up and down using a small motor to control the pressure inside the chamber.
The connection between the chamber and the reservoir passes through a large water tank
equipped with a temperature controller to set the temperature of the pressure chamber at the
required level. The actual pressure in the chamber is measured using an electronic pressure
transducer that can be connected to a personal computer. A wetting device is also used to
release a slow flow of saline solution on the cornea‘s anterior surface to prevent its drying.
The motor attached to the reservoir is connected to a variable-voltage power unit that can
change the motor speed within a wide range. The ability to change the motor‘s speed allows
the representation of both slow (static) and fast (dynamic) loading conditions to assess the
tissue‘s viscoelasticity. Creep tests can also be conducted by maintaining a constant level of
pressure for any reasonable duration of time. Corneal behaviour under loading is monitored
using electronic non-contact devices including a laser beam that measures corneal apical
deformation with 1 μm accuracy, and high-resolution digital cameras to record the changes
in cornea‘s profiles caused by pressure changes.
72 Ahmed Elsheikh
Figure 16. Overall views of inflation test rig showing the main components including (1) the pressure
chamber, (2) loading reservoir, (3) temperature controller, (4) laser beam, (5) digital cameras, and (6)
control computer.
Controlling the pressure application rate, and the tissue‘s hydration and temperature, and
the use of non-contact monitoring devices have led to experimental results with significantly
lower scatter compared to earlier studies which had less control of test conditions. The results
obtained using this rig are presented below and used to quantify some of the main
biomechanical properties of corneal tissue.
STRIP TEST SETUP

Strip testing is considerably simpler than inflation testing. Tissue strips are extracted
from corneal buttons in the desired anatomical direction, connected to mechanical clamps and
subjected to uniaxial tension loading in a material testing machine. However, care should be
taken to ensure that tissue is fresh and remains well preserved, specimen extraction is in the
right location and direction, and connection to the clamps is adequate. Corneal tissue can be
preserved in Optisol or Eusol C for up to 14 days post-mortem. Preservation, which also
maintains hydration, should continue during the tests as shown in Figure 17. Double blade
devices are needed to ensure the uniform width of specimens and the accurate positioning and
orientation of the specimen within the corneal button.
Current technology to connect tissue strips to mechanical clamps is inadequate. Too
much clamp tightening creates stress concentration in the specimen regions close to the
clamps, and can initiate premature failure. On the other hand, insufficient tightening allows
slippage and renders the results unreliable.
Figure 17. Corneal strip specimen (1) connected to mechanical clamps (2) in a uniaxial tension test.
74 Ahmed Elsheikh
The fact that the clamps provide non-uniform clamping effect across the thickness of
tissue (strong at the outer tissue layers and weaker at mid-thickness) is another cause for
concern. However, until a different technique to connect strip specimens to loading machines
is developed, mechanical clamps will have to be used with care to minimise the effect of the
above described deficiencies.
CORNEAL BEHAVIOUR CHARACTERISTICS

The following sections discuss the corneal behaviour as obtained experimentally using
the inflation and strip test rigs described above. In addition to human corneas, the discussion
will cover the behaviour of porcine corneas because of their reported biomechanical similarity
to, and suitability in some cases as test models for, human corneas. The behaviour
characteristics covered include the immediate hyperelasticity, hysteresis and anisotropy of
corneal tissue, and the viscoelastic properties represented in the strain-rate effects, and the
cornea‘s creep and stress-relaxation behaviour. Also included are the epithelium‘s
contribution to corneal biomechanics and the effects of the layered stromal construction.
Corneal Hyperelasticity
Under inflation testing, human corneas demonstrate clear hyperelastic (pressure-rise and
stress-strain) behaviour with initial low stiffness increasing gradually under higher pressure
levels, Figure 18. The end of the initial low stiffness stage coincides with posterior pressure
between 15 and 20 mmHg, which is close to the normal intraocular pressure range. The
average ratio of final stiffness (observed under IOP higher than 60 mmHg) to initial stiffness
(under 5-15 mmHg) has been found close to 11, 9 and 5 for corneas with donors within the
age ranges 50-64, 65-79 and 80-95 years, respectively [30].
200
180
Posterior pressure (mmHg)
160
140
120
100
80
60
40
20
0
0 0.1 0.2 0.3 0.4 0.5 0.6
Apical rise (mm)
Figure 18. Typical hyperelastic behaviour of human corneas under increasing posterior pressure.
180
160

140
120
100
80
60
40
20
0
0 0.1 0.2 0.3 0.4 0.5 0.6
Apical rise (mm)
Figure 19. Behaviour of porcine corneas under increasing posterior pressure.
Porcine corneas (aged 4-6 months) exhibit similar hyperelastic behaviour with a number
of important differences [35]. The change from the initial low stiffness stage to the final stage
with higher stiffness is more sudden and occurs around the lower pressure level of 10 mmHg,
Figure 19. The ratio between the final stiffness (under IOP higher than 60 mmHg) and the
initial stiffness (under 5-10 mmHg) values is about 5 on average.
Equations with different forms have been tried to fit the stress strain behaviour of human
corneas. The best fit was achieved with the exponential equation form ζ  A e  1
Bε
 
suggested earlier [7, 87] in which ζ and ε are the stress and strain, respectively, and A and B
are constants. A clear distinction has been found between the behaviour of specimens within
different age groups, and for this reason, three stress-strain equations have been developed of
the form:
ζ MPa  0.0070 e50 ε  1   Age 50-64 years (10)
ζ MPa  0.0077 e59 ε  1   Age 65-79 years (11)
ζ MPa  0.0083 e80 ε  1   Age 80-95 years (12)
The ages considered start from 50 years as this was the minimum age of corneal
specimens tested. Further work is therefore needed to develop stress-strain relationships for
younger human corneas.
The more sudden change in stiffness seen with porcine corneas made it difficult to find
an acceptable fit with the above exponential equation form, and an alternative third-order
polynomial form was adopted instead:
76 Ahmed Elsheikh
ζ(MPa) = - 350 ε3 + 56 ε2 – 0.21 ε (13)
The average porcine behaviour described by this equation is compared in Figure 20

against those of human corneas within the three age ranges considered. Overall, porcine
corneas appear to have lower initial and final stiffness values compared with all human
corneal groups.
0.12
0.1
0.08
Stress (MPa)
0.06
0.04 Human - 50-64 years

Human - 65-79 years
0.02 Human - 80-95 years
Porcine corneas
0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
Strain (mm/mm)
Figure 20. Comparison between the stress-strain behaviour of human and porcine corneas as obtained
experimentally using an inflation test rig.
Relationship between Intraocular Pressure and Young’s Modulus
The complex hyperelastic form of the cornea‘s stress-strain behaviour encouraged efforts
to find an easier, preferably linear, method to describe the behaviour. One such method is
based on presenting the behaviour as relationships between the applied posterior pressure, p,
(which represents the intraocular pressure, IOP) and the tangent Young‘s modulus, E. This
method was first attempted by Orssengo and Pye and resulted in the well-known linear
relationship [103]:
E = 0.0029 p (14)
Similar relationships have been derived from the experimental pressure-rise results
discussed above using a modification of Equation 2 of the form:
E tangent 
p  R 2
2 r t

1  ν  1  e-βη cos βη  (15)
where Δp and Δr are the increases in pressure and apical corneal rise within a particular
pressure increment, and E the instantaneous (tangent) Young‘s modulus [31].
Figure 21. Pressure-deformation and E-p behaviour of example human and porcine corneas, (a,b) 53
year old human cornea, (c,d) porcine cornea.
In spite of the hyperelastic form of both the pressure-rise and stress-strain results, the
relationship between Young‘s modulus, E, and posterior pressure, p, is close to linear. See for
example Figure 21 which shows the nonlinear pressure-rise behaviour of a 53 year old human
cornea and a porcine cornea, and the corresponding E-p relationships plotted for a pressure
range of 0-75 mmHg. In both cases, the E-p relationships are almost linear.
While the E-p results of human specimens are grouped according to the donor age, the
porcine results are treated as one group. Within each group, the results are used to find the
best fit with an E-p equation of the form: E = C . p + D, where C and D are constants. The
four equations obtained are:
E(MPa) = 0.0067 p(mmHg) + 0.1579 human corneas, age = 50-64 yrs (16)
78 Ahmed Elsheikh
E(MPa) = 0.0038 p(mmHg) + 0.1449 porcine corneas (19)
The four equations are compared in Figure 22, which shows porcine corneas with lower
E values compared to all human specimen groups. There is also a consistent stiffening effect
with age seen in human corneas.
Human corneas 50-64 years

1.4
Human corneas 65-79 years
1.2 Human corneas 80-95 years
Porcine corneas
1
E (MPa)
0.8
0.6
0.4
0.2
0
0 20 40 60 80
Figure 22. Comparison between E-p relationships for human and porcine corneas.
Corneal Hysteresis
Hysteresis is a viscoelastic property of corneal tissue characterised by the difference in

behaviour under loading and unloading conditions, Figure 23. The unloading behaviour has
been studied in a similar fashion to the loading behaviour and found to increase in stiffness
with age and pressure level. The stress-strain behaviour of the human corneal tissue under
unloading conditions can be described using the following equations [34]:
ζMPa   0.00004 e359 ε  1 Age 50-64 years (20)
ζMPa   0.00005 e427 ε  1 Age 65-79 years (21)
ζMPa   0.00025 e510 ε  1 Age 80-95 years (22)

Stress
Unloading
behaviour
Hysteresis
Loading
behaviour
Strain
Figure 23. Typical stress-strain behaviour of corneal tissue under both loading and unloading
conditions.
An earlier study has considered the unloading behaviour down from different posterior
pressure levels, and found the behaviour to be almost independent of the pressure at which
unloading starts [34], Figure 24. This particular feature makes the unloading behaviour easy
to characterise and implement in the numerical simulation of corneal response.
A direct measure of hysteresis can be obtained experimentally as the difference between
the areas under the loading and unloading stress-strain behaviour curves of corneal tissue.
Figure 25 depicts experimental results obtained earlier showing a consistent reduction in the
magnitude of hysteresis with corneal donor age [34]. The same figure shows a gradual
decrease in the areas under both loading and unloading stress-strain curves with age – a
finding which can be related to the higher stiffness (and hence lower maximum strain)
associated with increased age.
Figure 24. Average unloading behaviour of 80-95 year old corneas, (a) Unloading from three different
pressure levels plotted from an arbitrary starting corneal rise of 0.3mm, (b) The three patterns plotted
after shifting the behaviour lines with unloading from 55 and 110 mmHg.
80 Ahmed Elsheikh
Figure 25. Comparisons between areas under stress-strain loading and unloading curves for corneas
with different ages, (a) Sketches showing areas under behaviour curves, (b) Tests under fast pressure
application rate of 37.5 mmHg pressure per minute, (c) Tests under slow pressure application rate of
3.75 mmHg pressure per minute – Plots show error bars denoting ranges of variation between
specimens within the same group.
Corneal Anisotropy
Strip specimens extracted from human corneas have been tested to determine corneal
mechanical anisotropy [28]. The specimens were taken from the vertical, horizontal and
diagonal (superotemporal-inferonasal and superonasal-inferotemporal) directions and
comparisons of their stress-strain behaviour were held, Figure 26. Vertical specimens were
found to be slightly stiffer than horizontal specimens (by 8% on average) and significantly
stiffer than diagonal specimens (by 49% on average). This mechanical anisotropy is
compatible with the known preferential orientation of collagen fibrils of corneal stroma in the
vertical and horizontal directions described in Section 2.
Similar work on porcine corneas found little difference between vertical, horizontal and
the two diagonal directions [104]. The differences between the mechanical stiffness of strips
taken in the four directions were statistically insignificant and below 3% on average. This
apparent mechanical isotropy of behaviour was consistent with the microstructure of porcine
corneas in which stromal collagen fibrils were dominantly circumferential, and with no
preferential orientation in either the vertical, horizontal or diagonal directions.
Figure 26. Experimental assessment of the mechanical anisotropy of human donor corneas within
different age groups.
Epithelium’s Contribution to Corneal Biomechanics
The epithelium is the outermost layer of the cornea, comprising about 10% of its
thickness. It functions primarily to block the passage of foreign material into the eye, absorb
oxygen and nutrients from the tear film [1] and control the swelling of the stroma [65]. Little
work has been done to quantify the epithelium‘s contribution to corneal biomechanics,
82 Ahmed Elsheikh
although researchers have tended to discard the epithelium‘s contribution as insignificant [7,
92, 105].
Recent research used experimental testing of human corneas before and after removing
the epithelium and found that, on average, the epithelium was approximately 10% the
stiffness of stromal lamellae of the same thickness [38].
In order to compare the two sets of pressure-rise relationships (with and without the
epithelium), the two lines with the largest and smallest rises were identified within each set
and plotted together to mark the boundaries of behaviour seen with and without the
epithelium.
These lines were compared in Figure 27a, which shows little difference between the two
groups, although specimens with intact epithelium appeared to be slightly stiffer (i.e.
experiencing less deformation for the same pressure) than the other group. More precise
comparisons were obtained by considering the average stiffness values at specific pressure
levels between 10 and 100 mmHg. In this case, stiffness was defined as the posterior pressure
in mmHg required to produce an increase in corneal central rise of 1 mm. The stiffness
comparisons shown in Figure 27b demonstrate consistently higher average stiffness values
associated with the specimens with intact epithelium.
However, the ratios between corresponding values remained within 1-3% in all cases, and
these differences were not statistically significant (P>0.05 from t-tests) due to the
considerable overlap of the test pressure-rise results. With intact corneas exhibiting higher
stiffness by 1-3% than corneas without the epithelium, and with the epithelium being about
10% the corneal thickness, the stiffness of the epithelium is estimated between 9 and 27% that
of the stroma.
Figure 27. Comparison of pressure-rise behaviour between corneas with and without the epithelium –
Results of statistical analysis of differences between the two specimen groups are included on figure in
the form of t-test P values, (a) Range of pressure-rise results for all specimens tested, (b) Average
values of stiffness calculated at specific levels of posterior pressure.
Stromal Inter-Lamellar Adhesion
The corneal main layer, the stroma, is analogous to a stack of 300 sheets of copy paper
that have some ability to slide relative to each other if subjected to shear [9]. Lack of
knowledge of the shear behaviour affecting the relative movement of stromal sheets (or
lamellae) led to the exclusion of this feature from numerical simulations [106, 107]. Earlier
simulations concentrated on global behaviour, in which it was possible to assume
homogenous biomechanical properties (with no consideration of inter-lamellar shear) derived
from tests involving intact corneas. These modelling approaches would not be suitable for the
simulation of keratoconus for instance, with its abnormal inter-lamellar cross-links [108,
109], and refractive surgery procedures involving the severing of anterior lamellae.
A number of earlier studies determined through experimental means the adhesive tensile
strength of the connection between stromal lamellae [40-42]. Strip specimens extracted from
human stromal tissue were split at their ends to break the connection between two stromal
lamellae, Figure 28. The ends were then clamped and pulled apart, effectively subjecting the
interface between stromal lamellae to direct tension. This test, conducted on 68 human
specimens resulted in inter-lamellar tensile strength values between 0.069 and 0.077 MPa.
84 Ahmed Elsheikh
Figure 28. Testing of corneal tissue to determine the adhesive tensile strength of the connection
between stromal lamellae.
Further tests subjecting square specimens of stromal tissue to direct shear quantified the
shear behaviour of stromal tissue [39]. The behaviour was found to be nonlinear and had the
following exponential form:

η  0.009 e1.74   1  (23)
where  and  were the shear stress and shear strain, respectively. This result allowed an
indirect assessment of the assumption of stromal isotropy. If isotropy is maintained, the
behaviour patterns under shear and surface tension would have been interrelated. This
possibility is checked in a simple test, in which an approximate value of shear modulus, G, of
0.03 MPa was obtained from the shear stress-strain Equation 23 (0.02 MPa shear stress
divided by 0.65 strain). With an assumed value of Poisson‘s ratio, , of 0.49 (based on
considering corneal tissue an incompressible material [7]), the material‘s Young‘s modulus,
E, would be estimated as E = G [2 (1 + )] = 0.09 MPa. This value is considerably lower than
those reported in earlier experimental studies [30], indicating that the assumption of stromal
isotropy was invalid.
Corneal Long-Term Behaviour
Behaviour under long-term loading can be determined using two main tests, namely;
creep and stress-relaxation. In creep tests, specimens are kept for a period of time under a
constant level of load or stress while monitoring the gradual increase in deformation. On the
other hand, stress-relaxation tests involve subjecting specimens to a certain load, then
maintaining the resulting deformation for a period of time while monitoring the gradual
decline in stresses, Figure 29.
Figure 29. Overview of the stress and strain progression in (a,b) creep and (c,d) stress-relaxation tests.
The creep behaviour of human corneas attached to the inflation test rig and subjected to a
constant posterior pressure was found to differ with age as shown in Figure 30 and to follow
Equations 24-26 [35]:

ε creep  16  10 -6 e 2.28 log t - 1
1  age = 50-64 yrs (24)

ε creep  16  10 -6 e 2.23 log t - 1
1  age = 65-79 yrs (25)

ε creep  18  10 -6 e 2.11 log t - 1
1  age = 80-95 yrs (26)
where t is the creep time in seconds. A similar equation was developed for porcine corneas of
the form:

ε creep  5 10 -6 e 2.02 log t - 1
1  porcine corneas (27)
Figure 30 demonstrates a gradual decrease in the average human corneal creep with age,
which is more evident between the ages 50-64 and 65-79 years. However, there is a notable
overlap of results as demonstrated by the error bars in the figure. A statistical analysis of the
results using t-test shows that the association between increased age and reduced creep strain
is insignificant.
86 Ahmed Elsheikh
0.008 Human corneas with age 50-64 years

Human corneas with age 65-79 years
Human corneas with age 80-95 years
Porcine corneas
Creep strain (mm/mm)

0.006
0.004
0.002
0.000
1.0 1.5 2.0 2.5 3.0 3.5 4.0
Log time (seconds)
Figure 30. Comparison between average creep behaviour of porcine corneas and human corneas with
different ages.
In the mean time, porcine corneas undergo significantly more creep strain compared to all
human corneas. The ratio of average creep strain experienced by human and porcine corneas
at 300 and 1000 seconds is 49% and 42%, respectively. Figure 30 also shows that the overlap
between porcine and human creep-time results is limited to the first stages of the test, up to
about 100 seconds, beyond which there is clear separation. A statistical analysis of the two
sets of data further returned significant association between creep results and specimen type
with P < 0.001.
Stress-relaxation tests were also conducted, in which corneal strips were held with the
length reached under a uniaxial load of 4N, then 8N and 12N. The sustained load, and hence
the stress, were observed to reduce gradually with time. The stress reduction, or relaxation,
was highest at the start of the test and reduced gradually with time. The stress-time results
obtained in an earlier study were as shown in Figures 31a,c for human and porcine corneas,
respectively [35]. The results show a consistent trend in which porcine corneas were observed
to lose more of their initial stresses than human corneas. This trend was evident under all
three values of the initial load.
Dividing the stresses in each test by the initial stress value under either 4N, 8N or 12N,
and obtaining the average behaviour pattern in each case resulted in the plots in Figures
31b,d. Displaying the results in this form reveals two trends including an evidence that
porcine corneas relax quicker than human corneas. After 400, 800 and 1200 seconds, porcine
corneas respectively lose 49%, 55% and 59% of their initial stresses on average. The
corresponding average stress reductions in human corneas are 28%, 30% and 32%. The
differences in relaxation rates between human and porcine corneas are statistically significant
(P<0.001). Further, the stress-relaxation behaviour appears to be largely independent of the
initial applied stress. The error bars displayed in Figures 31b,d show significant overlap
between the results obtained under different initial applied loads.
Figure 31. Stress relaxation behaviour of human and porcine corneas, (a,b) Human corneal behaviour,
(c,d) Porcine corneal behaviour, (a,c) Stress-time experimental results, (b,d) Results plotted in relation
to initial stress values.
The average trend lines are also close and do not show any consistent trends related to the
value of initial load. This characteristic simplifies the development of stress-relaxation
behaviour patterns for use in numerical simulation applications.
Corneal Dependence on Strain-Rate
As a viscoelastic material, the behaviour of corneal tissue is time-dependent and is

expected to vary with the rate at which the material is strained or loaded. Experimental tests
involving vertical strips of human and porcine corneal tissue have been conducted under
uniaxial tension and with varying strain rates. With specimen length of 12 mm, width of 4
88 Ahmed Elsheikh
mm and elongation rates between 0.1 and 25 mm/minute, the strain rate varied between 0.008
and 2.1 per minute. As shown in Figure 32, the stiffening experienced with increasing the
strain rate was significant at the slow end of the rate range and became progressively less
notable as the strain rate increased. Under slow loading, increasing the strain rate from 0.008
to 0.08 per minute caused an increase in material stiffness of 47% and 9% on average in
human and porcine corneas, respectively. In contrast, the corresponding average increases in
stiffness that corresponded to increasing the strain rate from 0.83 to 2.1 were 1.5% and 1.2%.
The consistent difference in the response of human and porcine corneal tissue to strain rate
changes remains unexplained.
Figure 32. Effect of strain rate on the behaviour of (a) a human, and (b) a porcine corneal strip
specimen tested under uniaxial tension.
Concluding Remarks on Corneal Material Behaviour
Characterisation of the biomechanical properties of corneal tissue is key to developing

proper understanding of overall ocular behaviour, and response to disease, injury and surgery.
It is also essential for the construction of accurate predictive numerical simulations of ocular
behaviour, which have several clinical applications including improving the accuracy of
tonometry, the planning of refractive surgery procedures, the design of rigid contact lenses,
the treatment of myopia and keratoconus, and guiding surgical interventions following eye
injuries.
The experimental studies reported above led to fast progress in our understanding of the
human cornea‘s biomechanical behaviour, but several gaps still exist in our knowledge.
Overall hyperelastic and hysteretic behaviour, and how it changes with age, has been
characterised, but as the corneas tested were aged 50 years or over, similar studies are needed
to consider younger corneas. The epithelium‘s contribution to corneal biomechanics has been
quantified and further work is needed to assess the endothelium‘s and Bowman layer‘s
contributions, and to distinguish between the anterior and posterior parts of the stroma (which
differ in lamella interweaving). The correlation between corneal microstructure and
biomechanical behaviour has been established in normal healthy eyes, and similar
correlations are needed in swollen eyes (e.g. due to diurnal effects) and in eyes with
keratoconus. Further work is also needed to build similar knowledge in the biomechanical
behaviour and microstructure of the sclera, without which it becomes difficult to understand
overall ocular behaviour or to construct predictive numerical models of whole eye globes.
It is also important to have the ability to measure corneal and scleral biomechanical
properties in vivo. This is necessary if we are to build accurate and reliable patient-specific
ocular numerical simulations, especially with the notable variations in biomechanical
behaviour observed experimentally in same-age donor corneas. Recently, Reichert produced
the Ocular Response Analyzer (ORA), a tonometer that produces a hysteresis parameter and
claims that this parameter provides a measure of corneal biomechanics [110, 111]. However,
the true biomechanical origin of the hysteresis measurement, and which biomechanical
property it represents, is not yet fully understood.
Overall, and in spite of the current gaps in our knowledge, it remains true that the
significant progress made to date in corneal material characterisation, topography mapping
and microstructure detailing is offering an unprecedented opportunity to construct accurate
predictive numerical simulations of corneal response to disease, injury and surgery. This
subject will be discussed in some detail in the remainder of this chapter.
NUMERICAL MODELLING OF OCULAR BIOMECHANICS

In general, a numerical model is a computer-based model that describes the physical
behaviour of a structure and is able to predict its response to external loading. The structure is
divided into a large number of simple parts called elements, and the behaviour of each
element is described in a set of equilibrium equations normally presented in a matrix form.
The matrices are then assembled together to form overall equilibrium equations for the
structure. The mathematical assembly of individual matrices is done in the same way as the
90 Ahmed Elsheikh
elements can physically be put together to form the actual structure. Solving the overall
equilibrium equations under any set of given loads then provides a prediction of the
structure‘s response to the loads including its deformation and stress distribution [112]. This
modelling and solution process is commonly described as finite element analysis.
An example corneal model is shown in Figure 33. The model has four element layers
including anterior and posterior layers representing the epithelium and endothelium,
respectively, and two internal layers representing the stroma. Each layer is divided into 2904
solid prismatic elements, each with 5 faces and 6 corner points, called nodes. The elements
are arranged in 6 radial segments and 22 circumferential rings in such a way that ensures all
element internal angles remain between 20º and 70º. The model has a thickness that varies
from 540 µm at the centre to 690 µm along the edge, central anterior radius of 7.8 mm,
anterior shape factor, p = 0.82, and corneo-scleral diameter of 11.5 mm.
For finite element (FE) models to be accurate and predictive of real behaviour, the
models need to closely represent the structure‘s geometry, topography, boundary conditions,
loading regimes and material properties. In FE ocular models, these conditions include the
non-uniform corneal thickness, the elliptical topography, the connection to the sclera, the
uniform effect caused by the IOP, and the hyperelastic, hysteretic, anisotropic, viscoelastic
and layered behaviour of corneal tissue. However, since the introduction of these properties
can make the construction of FE models quite complex and the analysis costly and time
consuming, it becomes important to distinguish between the properties with significant effect
on model predictions and those that complicate the analysis without important improvements
in accuracy.
It is well known that all FE models involve simplifications in topography, material
properties, boundary conditions and/or loads [113-115].
Figure 33. Views of a corneal numerical model with 4 layers, each with 2,904 elements arranged in 6
segments and 22 element rings, (a) Top (anterior) surface of model showing model segments and
element rings, (b) View of bottom (posterior) surface and thickness of model, (c) Cross-sectional view
showing the 4 layers of the model representing the epithelium, stroma and endothelium.
For instance, the elliptical corneal topography is sometimes approximated as a spherical

surface, or the hyperelastic material behaviour idealised using a linear-elastic model. Some
simplifications are essential as they reduce the complexity of the model, lower its
development and execution costs and make it easier to modify and adapt. However, the more
simplifications adopted in FE model construction, the farther away from reality the model
becomes, eventually affecting its reliability. It is therefore important in constructing FE
models that the user adopts simplifications where they are likely to have a negligible effect
and avoid them when they start to erode the accuracy of predictions. This exercise requires
experience and validation against known experimental behaviour.
The following discussion aims to assess the effect of simplifications in FE model
construction with particular attention to simplifications in corneal topography, material
behaviour and boundary conditions.
The discussion refers to numerical results for a corneal model under two simple load
cases; a uniform posterior pressure (simulating the intraocular pressure, IOP) and a concentric
anterior point load, and the simplifications in a particular parameter are assessed by
comparing their effect on the simulation results. This exercise is followed by an overall
discussion on how to improve the accuracy of ocular numerical models and to allow them to
have a more useful role in clinical practice.
The two loading cases considered have been chosen for their simplicity and because they
represent idealised situations where the cornea is subjected to either uniform membrane
stresses under posterior pressure, or localised bending stresses under point loads. They have
been preferred to subjecting the cornea to more realistic loadings such as tonometric pressure
or external impacts since these loads create combinations of membrane and bending stresses,
which would make the discussion of results less straightforward. The analysis used the model
described above and shown in Figure 33, with some modifications added in the study of each
simplification as detailed below.
For simplicity, the following discussion focuses on the deformation recorded at the centre
point of the anterior surface, described as the corneal central, or apical, rise. Stiffness as used
in the discussion of results means corneal resistance to deformation, and in particular central
corneal rise.
SIMULATION OF CORNEAL NON-UNIFORM THICKNESS

The effect of corneal thickness variation is studied by considering two models; one with
variable thickness; from 540 µm at the centre to 690 µm along the edge, and one with
constant (average) thickness of 615 µm, Figure 34a. The results shown in Figures 34b, c
demonstrate a notable effect of thickness variation in models under point loads. The stiffness
increases by 10% on average when adopting a constant thickness of 615 µm, rather than
variable thickness. The effect under uniform posterior pressure is considerably lower, and
limited to 1.6%.
92 Ahmed Elsheikh
Figure 34. Behaviour predictions obtained using two models (a); one with variable corneal thickness
and one with constant thickness, under (b) a uniform posterior pressure, and (c) a concentric point load.
The structural effects of the two loading conditions differ considerably. In the first
loading condition, point loads create highly localised bending stresses around the load
application point, and this concentration of stress makes the behaviour dependent on the
thickness under the load, and almost independent of the thickness variation elsewhere. On the
other hand, the behaviour under the tension stresses caused by posterior pressure depends on
the thickness in all parts of the corneal surface, and in this case, replacing the real-life
variable thickness with an average constant thickness is not expected to lead to notable
change in overall corneal behaviour.
SIMULATION OF CORNEAL TOPOGRAPHY

The cornea has a prolate elliptical profile giving it maximum curvature (or smallest
radius) at the centre and flatter surface towards the limbus. The effect of elliptical profile is
studied under both uniform posterior pressure and anterior point load, and the results are
depicted in Figure 35. Models with shape factors, p, between 0.6 and 1.2 are considered to
cover the average reported prolate value, 0.8, the spherical case, 1.0, and an oblate case, 1.2
[43-45]. The results show a small, but notable, effect on models subjected to point loads (max
effect = 6.3%) and a more evident effect, exceeding 28%, under posterior pressure.
Figure 35. Effect of corneal shape factor, p, on predicted behaviour under (a) a uniform posterior
pressure, and (b) a concentric point load.
The highly localised stresses created by point loads are mainly dependent on the
thickness under the load, and corneal topography away from the load application point has
little effect. On the other hand, the stresses developed in the cornea due to posterior pressure
depend on the topography. When the corneal median surface matches the catenary
topography, the cornea experiences only tension membrane stresses (i.e. no bending stresses)
and hence undergoes small deformations. However, as the cornea deviates from the catenary
topography, bending stresses with increasing values arise leading to increased deformation
and reduced stiffness.
SIMULATION OF CORNEAL BOUNDARY CONDITIONS

The connection to the sclera presents a problem in corneal simulation as the sclera is not
stiff enough to prevent all deformation along corneal edge and is not flexible enough to allow
free edge movement.
This makes the use of pinned supports along corneal edge unrepresentative of real
conditions. One possible technique to avoid this problem is to use whole ocular models
(involving at least both the cornea and the sclera), but while this solution presents a close
representation of the cornea‘s natural conditions, it enlarges the model significantly and
makes it expensive to build and analyse.
Another solution is to provide the cornea with titled supports to consider the expected
sclera deformation, Figure 36a.
In this case, the slope of the supports is controlled such that the deformations of the
cornea-only model and the whole-globe model almost match. The results of this exercise are
demonstrated in Figure 36.
94 Ahmed Elsheikh
Figure 36. Behaviour of (a) cornea-only models with titled supports under (b) a uniform posterior
pressure, and (c) a concentric point load.
The results show first that the support tilts suitable for simulations under posterior
pressure and point load were 53º and 60º, respectively. Second, as the behaviour under point
loads is dependent mainly on the thickness under the loads and largely insensitive to model
changes elsewhere, it is reasonable to expect the model‘s low sensitivity to variations in
support tilt in this case. The results also show the high sensitivity of analysis under posterior
pressure to the support tilt. This finding should be considered carefully as it can make cornea-
only models in this case less attractive, and much less reliable, than whole eye models.
SIMULATION OF CORNEAL HYPERELASTIC BEHAVIOUR

Corneal tissue is known to possess hyperelastic behaviour with initial low stiffness,
gradually increasing under higher stress levels. Considering this behaviour pattern is a
challenge since the engineering materials commonly considered in commercial FE packages
have the opposite behaviour, i.e. initial high stiffness, decreasing under higher stresses. For
this reason, several earlier modelling studies simplified the corneal behaviour into a linear-
elastic pattern, and this was justified by the assumption that common loads on the cornea
create low stresses. In Figure 37, this assumption is tested under both posterior pressure and a
central point load, and in both cases, two models were considered with hyperelastic and
linear-elastic material behaviour patterns, respectively. The hyperelastic model considered the
following stress-strain relationship developed earlier for corneas aged 50-65 years [30]:

ζ MPa  0.0070 e50 ε  1 ,  (28)
while the linear-elastic model had a constant Young‘s modulus, E, equal to the E of the
hyperelastic model under low strain, ε = 0.01 (E = 0.56 MPa). The results show that the linear
elastic model provided reasonable results under low load levels up to about 0.002 MPa (15
mmHg) or 0.01 N, beyond which the deviation between the two models became unacceptably
high.
Figure 37. Effect of adopting a linear-elastic material behaviour pattern on the model‘s response to (a) a
uniform posterior pressure, and (b) a concentric point load.
SIMULATION OF CORNEAL LAYERED CONSTRUCTION

The cornea‘s three main layers, the epithelium, stroma and endothelium, have different
material properties. The stroma, with its fibrous structure, has the highest stiffness, while the
cellular epithelium and endothelium are much more flexible. An earlier study estimated the
epithelial stiffness at 10% that of the stroma [38]. No study has been carried out on
endothelial biomechanics, but its cellular structure could justify assuming it had similar
properties to the epithelium. The effect of considering the distinctive material properties of
each layer on corneal behaviour was assessed and the results are shown in Figure 38.
96 Ahmed Elsheikh
Figure 38. Effect of considering the distinctive material properties of the main corneal layers on
predicting the behaviour under (a) a uniform posterior pressure, and (b) a concentric point load.
The homogenous model had a uniform stiffness equivalent to that of the heterogeneous
model under loads creating membrane tension. This condition meant that:
(E . t)homogenous = (E . t)epithelium + (E . t)stroma + (E . t)endothelium (29)
In this relationship, corneal thickness, t, was split into 8%, 90% and 2% parts for the
three layers, respectively [64]. Eepithelium and Eendothelium were also both taken as 10% of Estroma.
The results show almost negligible effect of varying the material properties on the behaviour
under posterior pressure, as would be expected. However, the effect was clear under a
concentric point load with an average reduction in stiffness of 15%.
Since posterior pressure creates mainly membrane tension stresses, these stresses will be
shared by the cornea‘s three layers according to their respective membrane stiffness values
(E.t). However, as the overall membrane stiffness of the cornea does not change by varying
the distribution of stiffness between layers (according to Equation 29), the overall behaviour
under posterior pressure is not expected to change.
On the other hand, the bending stresses caused by point loads depend on the cornea‘s
bending stiffness calculated as the sum of Ei . Ii for all corneal layers, where Ii (the second
moment of area of layer i) is ti3 / 12 + ti  zi2, and zi the distance between the mid-thickness of
layer i and the cornea mid-thickness. Therefore, placing the layers with high Young‘s
modulus, E, nearer the cornea mid-thickness (hence lower zi), as in the second model, reduces
the overall bending stiffness and leads to increased deformation.
SIMULATION OF STROMAL INTER-LAMELLAR ADHESION

The cornea has weak shear stiffness at the surfaces separating stromal lamellae [9]. As
was discussed above, the stromal inter-lamellar adhesion was assessed experimentally, and
 
the shear stress-strain, , relationship η MPa   0.009 e1.74   1 was derived [39]. The
effect of stromal adhesion on corneal behaviour was considered by reducing the stromal inter-
layer contact in the numerical simulations from full to the level observed experimentally. The
results depicted in Figure 39 show a consistent trend in which corneal stiffness degraded with
reducing stromal inter-lamellar interaction. On average, the stiffness reduced by 6% and 56%
for the cases under posterior pressure and point load, respectively.
Figure 39. Effect of stromal inter-laminar adhesion on corneal behaviour under (a) a uniform posterior
pressure, and (b) a concentric point load.
Since posterior pressure is transferred through membrane tension, which depends on the
thickness and stiffness of individual layers, and not on inter-layer contact, reducing the level
of contact does not have a considerable effect on overall behaviour under pressure [116]. On
the other hand, reducing inter-layer contact reduces the second moment of area (as the
stiffness term ti  zi2 is to be multiplied by a reduction factor which depends on the level of
interaction) and therefore affects the resistance to deformation under bending actions such as
those created by point loads.
SIMULATION OF CORNEAL ANISOTROPY

The human cornea has a preferential orientation of stromal collagen fibrils in the vertical
(superior-inferior, SI) and horizontal (temporal-nasal, TN) directions, leading to anisotropy in
biomechanical behaviour [24, 26]. The degree by which the SI and TN directions exhibit
higher stiffness than diagonal directions has been determined earlier [28] and discussed
above. Implementing anisotropy in corneal simulations was studied by considering two multi-
layer numerical models; one with anisotropic and another with isotropic behaviour,
respectively. The models in this study had 8 element layers representing the stroma, instead of
the two layers used in other parts of the study.
98 Ahmed Elsheikh
Figure 40. Effect of introducing material anisotropy on predicted behaviour under (a) a uniform
posterior pressure, and (b) a concentric point load.
In the first model, odd-numbered stroma layers were assumed to have fibrils in only the
vertical direction while even-numbered layers had fibrils in the horizontal direction. The
layers were given a high stiffness in the fibril direction and 10% this value in the
perpendicular direction.
In the second isotropic model, the stiffness was the same in all directions and taken as the
average of the stiffness values in the two directions of the first model. The models were used
to predict the behaviour under a posterior pressure and a concentric point load and the results
showed little effect, below 2% under posterior pressure, and 4% under point load, caused by
introducing anisotropy, see Figure 40.
SIMULATION OF CORNEAL HYSTERESIS

Considering corneal hysteresis in the material definition in FE models is perhaps the most
difficult and challenging feature in model construction. This feature requires that the corneal
material is defined with both loading and unloading behaviour patterns, and that the
simulation can select the pattern that fits the stress history of each element. Following each
change in loading, the simulation calculates the stress level within each element and
determines whether the element is to undergo further loading or unloading, and based on this
decision, the element is allowed to follow the more suitable behaviour pattern from the stress
level determined earlier, Figure 41. This process is difficult to program and makes the
analysis process quite complex.
Figure 41. (a) Behaviour of corneal tissue under both loading and unloading conditions, and the stress
development in a particular element during an FE analysis including (b) the stress history, and the
behaviour adopted if the element is to undergo either (c) further loading or (d) unloading.
Nevertheless, an assessment is needed of whether hysteresis should be incorporated in

material definition in the first place. The first trials under monotonically increasing posterior
pressure or point loads showed no effect of considering hysteresis. Since all elements
underwent further loading in all analysis steps, consideration of the unloading behaviour was
not required.
However, further tests showed different outcomes. First, a simple corneal model was
subjected to a posterior pressure that increased gradually to 30 mmHg then decreased down to
0. Ignoring hysteresis resulted in pressure-rise predictions that were the same under loading
and unloading, Figure 42a. However, when hysteresis was considered, the unloading
behaviour differed from the loading behaviour as would be expected in experimental tests,
Figure 42b.
Another test involved the simulation of the Ocular Response Analyzer (ORA) – a
tonometric technique that applies an air pulse on the cornea in increasing then decreasing
intensity. Figures 43b-g show the stress distributions on a corneal cross-section at different
stages of loading; first by intraocular pressure and then by external air pulse. The figures
illustrate the unloading experienced in the middle region of the model during the second stage
(force-out) of the ORA procedure. It is therefore hypothesised that this region should follow
the material unloading behaviour in order to trace corneal response during the full ORA
procedure.
100 Ahmed Elsheikh
Figure 42. Model predictions of the behaviour of a cornea subjected to a cycle of pressure increase and
decrease while (a) ignoring and (b) considering the unloading behaviour of corneal tissue.
St re s s i n MP a
(a) (b)
All are as under
increased loading
a
rne
Co
u re
ss
r pre
(b) ula (c)
oc
ra
In t
Tonom etric pressure Tonom etric pressure

(P 1) Area un dergoing (Pm ax) All are as under
u nloading increased loading
Area un dergoing
further loading
re
u re ssu
e ss r pre
r pr ul a
ul a oc
oc
tra
(d) (e)
tra
In
In
Area un dergoing
Tonom etric pressure further loading
(P 2) Area un dergoing
All are as under u nloading
u nloading
u re
ss
su r
e
r pre
re s u la
rp
oc
ul a
ra
o c
In t
(f) (g)
tra
In
Figure 43. Stress distribution diagrams for a cornea with CCT = 540 µm, PCT = 690 µm, R = 7.8 mm,
p = 0.82, age = 57 years and IOP = 15 mmHg, during the progress of the ORA procedure showing areas
of loading and unloading, (a) Key showing stress values in MPa, (b) Cross-sectional view at start, (c)
Stresses after applying the intraocular pressure on the model‘s posterior side, (d) Stresses at first
applanation, (e) Stresses under Pmax, beyond which the air pressure starts to decrease, (f) Stresses at
second applanation, (g) Stresses after complete removal of air pressure.
In order to test this hypothesis, the ORA procedure was simulated using two models that
considered and ignored corneal hysteresis, respectively. The models were subjected first to an
intraocular pressure of 15 mmHg. The ORA air pressure was then gradually applied until
applanation was detected at P1 (the first applanation pressure) of 21.3 mmHg. The pressure
was further increased to a peak value, Pmax, of 53 mmHg according to the relation: Pmax =
1.1713 P1 + 28.106 [117]. After reaching Pmax, the pressure was decreased gradually while
looking for the second applanation pressure, P2.
The deformation was monitored at two points; the corneal centre and a point 0.6 mm
away from the centre (the next point on the model‘s anterior surface away from the centre).
Applanation was detected when the two points had the same coordinate in the anterior-
posterior direction.
Figure 44 shows that when hysteresis was considered, the deformation under unloading
differed from that under loading, as would be expected. As a result, the applanation pressures,
P1 and P2, at which the central cornea flattened under air pressure, were different. However,
when hysteresis was omitted, the two behaviour patterns were identical, and subsequently P2
was incorrectly equal to P1.
A similar test involving the simulation of the Goldmann Applanation Tonometer (GAT)
was conducted. Unlike ORA, GAT is a contact tonometer that pushes a solid cylindrical
device against the central cornea until applanation over a circular area with 3.06 mm diameter
is achieved.
The behaviour was simulated numerically (including the initial application of intraocular
pressure followed by contact with the tonometer), and the stress distribution during the two
loading stages is depicted in Figure 45.
7.8
Simulation that
P1 = 21.3 mmHg P2 = 18.3 mmHg
considers hysteresis
7.6
7.4
Simulation that does not

consider hysteresis
Corneal rise (mm)
7.6
7.4
7.2
P1 = 21.3 mmHg P2 = 21.3 mmHg
7 Centre point
Point 0.6mm away from centre
6.8
Corneal shape
Inflation step up to 15 mmHg Increasing air pressure Decreasing air pressure

up to Pmax = 53 mmHg down to 0 mmHg
Figure 44. Corneal rise as measured at the centre and a point 0.6mm away from the centre during
inflation under 15 mmHg followed by increasing then decreasing air pressure as applied by the Ocular
Response Analyzer.
102 Ahmed Elsheikh
Figure 45. Stress development under IOP and GAT tonometry pressure, (a) Cornea under IOP with all
areas following loading behaviour, (b) Cornea under both IOP and tonometry pressure with area under
tonometer experiencing unloading.
The figure shows how the central corneal region undergoes unloading during contact with
the tonometer – a region which should follow the material‘s unloading behaviour in order for
the simulation to provide an accurate representation of the GAT procedure.
An attempt has been made to assess the importance of considering corneal hysteresis in
the simulation of GAT. Two cornea models have been developed with hysteresis considered
in one and ignored in the other. Similar to ORA simulations, both models have CCT = 540
µm, PCT = 690 µm, R = 7.8 mm, p = 0.82, age = 57 years and IOP = 15 mmHg. While the
first model, which considers hysteresis, estimated a GAT reading of IOP (denoted IOPG) of
15.19 mmHg, the second model which ignores hysteresis predicted an IOPG value of 13.96
mmHg. The considerable difference (8%) between the two IOPG values indicate the
importance of considering hysteresis although the corneal zone expected to undergo
unloading is small in this case.
CONCLUSION
Numerical simulation has a strong potential in corneal biomechanics applications. It can
be, and has been, used to improve the accuracy of tonometry [118, 119], the planning of
refractive surgery [5,6, 14, 107], the evaluation of damage caused by impacts [11], the design
of contact lenses, and the estimation of disease progression. This strong potential can be
realised with growing interest from various research groups [120, 121], fast advances in
numerical modelling, and significant contributions to knowledge and understanding of
corneal microstructure, topography and material characterisation.
However, a number of challenges currently exist and slow the progress of corneal
numerical simulation. First, the level of complexity of corneal microstructure with variable
interweaving of stromal lamellae, and change in preferential orientation of collagen fibrils
from vertical and horizontal at the centre to circumferential near the limbus, make the
development of a corneal model that is truly representative of real-life conditions highly

complex.
Second, the material properties of corneal tissue pose serious challenges in simulation.
On a basic level, the material‘s hyperelasticity and hysteresis can be considered, albeit with
some difficulty, in numerical simulation. However, these properties are not constant, but vary
with age, rate of loading and anatomical direction, and therefore a material model which is
adequate for a particular case, could be entirely inaccurate in another. One typical example is
the difference in speed of loading experienced under contact and non-contact tonometry,
which would require widely different material models [118, 119]. The stiffening effect of
ageing adds another complication to the behaviour under both loading and unloading [30, 34].
Third and most important is the current lack of understanding of some fundamental
corneal properties. These include the effect of swelling (following sleep or contact lens wear)
on corneal microstructure and biomechanical behaviour, the changes in microstructure
associated with the progression of diseases such as keratoconus, the effects of long term
exposure to high levels of IOP in glaucoma, and the wound healing process following surgery
and whether stromal tissue is able to regain its mechanical integrity after wound healing.
These effects must be quantified experimentally and/or clinically before they can be
accurately considered in numerical simulation models.
Fourth, the current level of understanding of other ocular components, most notably the
sclera, is not adequate to enable their proper inclusion in ocular simulations. While a few
studies conducted over the last 40 years helped improve understanding of scleral thickness
variation, regional biomechanical performance and microstructure [122-130], work is still
required to quantify the effect of sclera‘s hyperelasticity, hysteresis, viscoelasticity,
anisotropy, layered composition, microstructure, and medical conditions such as myopia on
scleral biomechanics. While this work will be important in its own right, it will enable the
construction of more reliable whole eye globe models without having to approximate the
boundary conditions of the cornea.
The above discussion is intended to point at the areas where corneal biomechanics
research is most needed, and to draw attention to the challenges slowing the development and
application of corneal numerical simulation. However, it must be emphasised that while work
to address these challenges is needed, corneal numerical simulation should not wait and
should progress based on current knowledge. What has been achieved over the past few
decades in corneal material characterisation, topography mapping and microstructure
detailing, is significant and sufficient for several important applications. Normal, healthy and
intact corneas can now be reliably modelled under actions such as contact and non-contact
tonometry or external impacts. These applications not only address important and relevant
clinical problems, they also help demonstrate the potential value of numerical simulation in
corneal biomechanics applications. Nevertheless, care should be taken not to expand
simulation into areas where knowledge of material behaviour or microstructure has not yet
been generated. These attempts would be likely to produce unreliable results, and might affect
the general perception of what numerical simulation can do in corneal applications.
Detailed clinical validation is also required to assess the accuracy of numerical
simulations and demonstrate their suitability in providing solutions to clinical problems.
Clinical validation will also help overcome the traditional reluctance of ophthalmologists and
optometrists to embrace predictive tools based on numerical simulations, and prepare the
ground for their wide acceptance.
104 Ahmed Elsheikh
REFERENCES
[1] Fatt I. Physiology of the eye: An introduction to the vegetative functions, Butterworths,
London, 1978.
[2] Roberts C. The cornea is not a piece of plastic. Journal of Refractive Surgery.
2000;16:407-413.
[3] Ethier CR, Johnson M, Ruberti J. Ocular biomechanics and biotransport. Annual
Review of Biomedical Engineering. 2004;6:249-273.
[4] Michelacci YM. Collagens and proteoglycans of the corneal extracellular matrix.
Brazilian Journal of Medical and Biological Research. 2003;36(8):1037-1046.
[5] Vito RP, Shin TJ, McCarey BE. A Mechanical Model of the Cornea: The Effects of
Physiological and Surgical Factors on Radial Keratotomy Surgery. Refractive Corneal
Surgery. 1989;5(2):82-88.
[6] Pinsky P, Datye V. A microstructurally-based finite element model of the incised
human cornea. Journal of Biomechanics. 1991;10:907-922.
[7] Bryant M, McDonnell P. Constitutive laws for biomechanical modeling of refractive
surgery. Journal of Biomechanical Engineering. 1996;118:473-481.
[8] Buzard KA. Introduction to bio-mechanics of the cornea. Refractive and Corneal
Surgery. 1992;8:127-138.
[9] Dupps WJ. Biomechanical modeling of corneal ectasia. J Refractive Surgery.
2005;21(2):186-190.
[10] Velinsky S, Bryant M. On the Computer-Aided and Optimal Design of Keratorefractive
Surgery. Refractive Corneal Surgery. 1992;8:173-183.
[11] Uchio E, Ohno S, Kudoh J, Aoki K, Kisielewicz LT. Simulation model of an eyeball
based on finite element analysis on a supercomputer. Br. J. Ophthalmology.
1999;83:1106-1111.
[12] Fernández DC, Niazy AM, Kurtz RM, Djotyan GP, Juhasz T. Finite element analysis
applied to cornea reshaping. Journal of Biomedical Optics. 2005;10(6):064018-1-11.
[13] Fernández DC, Niazy AM, Kurtz RM, Djotyan GP, Juhasz T. Computational modeling
of corneal refractive surgery. Proceedings of SPIE. 2004;Vol. 5314. Bellingham, WA.
[14] Alastrué V, Calvo B, Peña E, Doblaré M. Biomechanical modeling of refractive corneal
surgery. Biomechanical Engineering. 2006;128;150-160.
[15] Hanna, K., Jouve, F., Waring, G., and Ciarlet, P., 1992, ―Computer simulation of
arcuate keratotomy for astigmatism,‖ Refract. Corneal Surg. 8 _1992_ 152-163.
[16] Pinsky, P., and Datye, V., 1991, ―A microstructurally-based finite element model of the
incised human cornea,‖ J. Biomech. 10, pp. 907-922.
[17] Howland, H., Rand, R., and Lubkin, S., 1992, ―A Thin-Shell Model of the Cornea and
its Application to Corneal Surgery,‖ Refract. Corneal Surg. 8, pp.183-186.
[18] Bryant, M. R., Machi, V., and Juhasz, T., 2000, ―Mathematical Model of Picosecond
Laser Keratomileusis For High Myopia,‖ ASME J. Biomech. Eng. 16, pp. 155-162.
[19] Djotyan, G. P., Kurtz, R. M., Cabrera, D., and Juhasz, T., 2001, ―An Analytically
Solvable Model For Biomechanical Response of the Cornea to Refractive Surgery,‖ J.
Refract. Surg. 123, pp. 440-445.
[20] Huang, D., Tang, M., and Shekhar, R., 2003, ―Mathematical Model of the Corneal
Surface Smoothing After Laser Refractive Surgery,‖ Am. J. Ophthalmol. 135, pp. 267-
278.
[21] Yeh, H.-L., Huang, T., and Schachar, R., 2000, ―A Closed Shell Structured Eyeball
Model With Application to Radial Keratotomy,‖ J. Biomech. Eng. 122, pp. 504–510.
[22] Hanna, K. D., Jouve, F., Bercovier, M. H., and Waring, G. O., 1989, ―Computer
Simulation of Lamellar Keratectomy and Laser Myopic Keratomileusis,‖ J. Refract.
Surg. 4, pp. 222–231.
[23] Hanna, K. D., Jouve, F., Ciarlet, P., and Waring, G. O., 1989, ―Computer Simulation of
Arcuate and Radial Incisions Involving the Corneoscleral Limbus,‖ Eye 3, pp. 227–239.
[24] Boote C, Dennis S, Huang Y, Quantock AJ, Meek KM. Lamellar orientation in human
cornea in relation to mechanical properties. Journal of Structural Biology. 2005;149:1-
6.
[25] Müller LJ, Pelsa E, Schurmans LRHM, Vrensen GFJM. A new three-dimensional
model of the organization of proteoglycans and collagen fibrils in the human corneal
stroma. Experimental Eye Research. 2004;78:493-501.
[26] Meek KM, Newton RH. Organization of collagen fibrils in the corneal stroma in
relation to mechanical properties and surgical practice. Journal of Refractive Surgery.
1999;15(6):695-699.
[27] Aghamohammadzadeh H, Newton RH, Meek KM. X-ray scattering used to map the
preferred collagen orientation in the human cornea and limbus. Structure. 2004;12:249-
256.
[28] Elsheikh A, Brown M, Alhasso D, Rama P, Campanelli M, Garway-Heath DF.
Experimental assessment of corneal anisotropy. Journal of Refractive Surgery.
2008;24(2):178-187.
[29] Jayasuriya AC, Ghosh S, Scheinbeim JI, Lubkin V, Bennett G, Kramer P. A study of
piezoelectric and mechanical anisotropies of the human Cornea. Biosens Bioelectron.
2003;18:381-387.
[30] Elsheikh A, Wang D, Rama P, Campanelli M, Pye D. Assessment of corneal
biomechanical properties and their variation with age. Current Eye Research.
2007;32:11-19.
[31] Elsheikh A, Wang D, Pye D. Determination of the modulus of elasticity of the human
cornea. Journal of Refractive Surgery. 2007;23(8):808-818.
[32] Tallec PL, Rahier C, Kaiss A. Three-dimensional incompressible viscoelasticity in large
strains: Formulation and numerical approximation. Comput. Methods Appl. Mech. Eng.
1993;109:233-258.
[33] Boyce B, Jones R, Nguyen T, Grazier J. Stress-controlled viscoelastic tensile response
of bovine cornea. Journal of Biomechanics. 2007; 40(11):2367-2376.
[34] Elsheikh A, Wang D, Rama P, Campanelli M, Garway-Heath DF. Experimental
assessment of human corneal hysteresis. Current Eye Research. 2008;33:205-213.
[35] Elsheikh A, Alhasso D, Rama P. Biomechanical properties of human and porcine
corneas. Experimental Eye Research. 2008;86:783-790.
[36] Kampmeier J, Radt B, Birngruber R, Brinkmann R. Thermal and biomechanical
parameters of porcine cornea. Cornea. 2000;19:355-362.
106 Ahmed Elsheikh
[37] Scherer K, Eggert H, Guth H, Stiller P. Application of simulation techniques in human

eye corneal surgery. 13th International Conference of Society for Medical Innovation
and Technology. Berlin. 2001.
[38] Elsheikh A, Alhasso D, Rama P. Assessment of the epithelium‘s contribution to corneal
biomechanics. Experimental Eye Research. 2008;86:445-451.
[39] Elsheikh A, Ross S, Alhasso D, Rama P. Numerical study of the effect of corneal
layered structure on ocular biomechanics. Current Eye Research. In Press
[40] Schmack I, Dawson DG, McCarey BE, Waring BO, Grossniklaus HE, Edelhauser HF.
Cohesive tensile strength of human LASIK wounds with histologic, ultrastructural and
clinical correlations. Journal of Refractive Surgery. 2005;21:433-445.
[41] Smolek MK, McCarey BE. Interlamellar adhesive strength in human eyebank corneas.
IOVS. 1990;31(6):1087-1095.
[42] Smolek MK. Interlamellar cohesive strength in the vertical meridian of human eye bank
corneas. IOVS. 1993;34:2962-2969.
[43] Dubbelman M, Weeber HA, van der Heijde RGL, Völker-Dieben HJ. Radius and
asphericity of the posterior corneal surface determined by corrected Scheimpflug
photography. Acta Ophthalmol. Scand. 2002;80:379-383.
[44] Read SA, Collins MJ, Carney LG, Franklin RJ. The topography of the central and
peripheral cornea. IOVS. 2006;47:1404-1415.
[45] Guillon M, Lydon PM, Wilson C. Corneal topography: a clinical model. Ophthalmic.
Physiol. Opt. 1986;6:47-56.
[46] Aghaian E, Choe JE, Lin S, Stamper RL. Central Corneal Thickness of Caucasians,
Chinese, Hispanics, Filipinos, African Americans, and Japanese in a Glaucoma Clinic.
Ophthalmology. 2004;111:2211-2219.
[47] Cho P, Cheung SW. Central and peripheral corneal thickness measured with the
TOPCON specular microscope SP-2000P. Current Eye Research. 2000;21(4):799-807.
[48] Lam AKC, Chan JS. Corneal thickness at different reference points from Orbscan II
system. Clinical and Experimental Optometry. 2003;86(4):230-234.
[49] Ethier CR, JohnsonM, Ruberti J. Ocular biomechanics and biotransport. Annu. Rev.
Biomed. Eng. 2004. 6:249-73.
[50] Pouliquen Y. 1985. Fine structure of the corneal stroma. Cornea 3:168-77.
[51] Parry DA, Craig AS. Electron microscope evidence for an 80 A° unit in collagen fibrils.
Nature. 1979;282:213–215.
[52] Komai Y, Ushiki T. The three-dimensional organization of collagen fibrils in the
human cornea and sclera. Invest. Ophthalmol. Vis. Sci. 1991;32:2244-2258.
[53] Maurice DM, Monroe F. Cohesive strength of corneal lamellae. Exp. Eye Res.
1990;50:59.
[54] Kokott W. Über mechanisch-funktionelle Strukturen des Auges. Albrecht. v Grafes
Arch. Ophthalmol 1938;118:424-485.
[55] Meek KM, Blamires T, Elliot GF, Gyi TJ, Nave C. The organization of collagen fibrils
in the human corneal stroma: a synchroton x-ray distraction study. Current Eye
Research, 1987;6:841-846.
[56] Newton RH, Meek KM. The integration of the corneal and limbal fibrils in the human
eye. Biophysical Journal, 1998;75:2508-2512.
[57] Newton RH, Meek KM. Circumcorneal annulus of collagen fibrils in the human limbus.
Investigative Ophthalmology and Visual Science. 1998;39:1125-1134.
[58] Daxer A, Fratzl P. Collagen fibril orientation in the human corneal stroma and its
implication in keratoconus. Investigative Ophthalmology and Visual Science,
1997;38(1):121-1293.
[59] Hayes, S., Boote, C., Lewis, J., Sheppard, J., Abahussin, M., Quantock, A.J., Purslow,
C., Votruba, M., Meek, K.M., 2007. Comparative study of fibrillar collagen
arrangement in the corneas of primates and other mammals. The Anatomical Record
290:1542-1550.
[60] Boote C, Dennis S, Meek KM. Spatial mapping of collagen fibril organisation in
primate cornea—an X-ray diffraction investigation. Journal of Structural Biology 146
(2004) 359-367.
[61] Prangen, A., 1928. A study of the comparative anatomy of the extraocular muscles.
Trans. Am. Ophthalmol. Soc. 26:353-380.
[62] Green K (1991) Corneal endothelial structure and function under normal and toxic
conditions. Cell Biology Review 25, 169-207.
[63] Johnson M, Shapiro A, Ethier CR, Kamm RD. Modulation of outflow resistance by the
pores of the inner wall endothelium. IOVS. 1992;33:1670-1675.
[64] Coster DJ. Cornea. BMJ Publishing Group, London. 2001.
[65] Patel, S., Alio, J.L., Perez-Santonja, J.J., 2004. Refractive index change in bovine and
human corneal stroma before and after LASIK: A study of untreated and re-treated
corneas implicating stromal hydration. IOVS. 45(10), 3523-3530.
[66] Beuerman RW, Pedroza L (1996) Ultrastructure of human cornea. Microscopy
Research and Technique 33, 320-335.
[67] Danielsen CC. Tensile mechanical and creep properties of Descemet‘s membrane and
lens capsule. Experimental Eye Research 79 (2004) 343-350.
[68] Merindano MD, Costa J, Canals M, Potau JM, Ruano D. A comparative study of
Bowman‘s layer in some mammals: Relationships with other constituent corneal
structures. Eur. J. Anat., 6 (3) : 133-139 (2002)
[69] Seiler T, Matallana M, Sendler S, Bende T. Does Bowman's layer determine the
biomechanical properties of the cornea? Refract Corneal Surg. 1992;8(2):139-42.
[70] Feltgen, N., Leifert, D., Funk, J., 2001, ―Correlation between central corneal thickness,
applanation tonometry and direct intracameral IOP readings,‖ British Journal of
Ophthalmology, 85, pp. 85-87.
[71] Laiquzzaman, M., Bhojwani, R., Cunliffe, I., Shah, S., 2006, ―Diurnal variation of
ocular hysteresis in normal subjects: relevance in clinical context,‖ Clinical and
Experimental Ophthalmology, 34, pp. 114-118.
[72] Foster, P. J., Machin, D., Wong, T.-Y., Ng, T.-P., Kirwan, J. F., Johnson, G. J., Khaw,
P. T., Seah, S. K. L., 2003, ―Determinants of intraocular pressure and its association
with glaucomatous optic neuropathy in Chinese Singaporeans: The Tanjong Pagar
study,‖ IOVS, 44, pp. 3885-3891.
[73] Lleó A, Marcos A, Calatayud M, Alonso L, Rahhal SM, Sanchis-Gimeno JA. The
relationship between central corneal thickness and Goldmann applanation tonometry.
Clin. Exp. Optom. 2003;86(2):104-108.
[74] Shah, S., Spedding, C., Bhojwani, R., Kwartz, J., Henson, D., McLeod, D., 2000,
―Assessment of the diurnal variation in central corneal thickness and intraocular
pressure for patients with suspected glaucoma,‖ Ophthalmology, 107(6), pp. 1191-
1193.
108 Ahmed Elsheikh
[75] Hirji, N. K., Larke, J. R., 1978, ―Thickness of human cornea measured by topographic
tachometry,‖ American Journal of Optometry, 55, pp. 97-100.
[76] Martola, E. L., Baum, J. L., 1968, ―Central and peripheral corneal thickness,‖ Arch.
Opththalmol, 79, pp. 28-30.
[77] Maurice, D. M., Giardini, A. A., 1951, ―A simple optical apparatus for measuring the
corneal thickness and the average thickness of the human cornea,‖ British Journal of
Ophthalmology, 35, pp. 169-177.
[78] Hennighausen H, Feldman ST, Bille JF, McCulloch AD. Anterior-posterior strain
variation in normally hydrated and swollen rabbit cornea. IOVS. 1998;39:253-262.
[79] Nash IS, Greene PR, Foster CS. Comparison of mechanical properties of keratoconus
and normal corneas. Exp. Eye Res. 1982; 35:413-24.
[80] Fung, Y.C. Biomechanics: Mechanical properties of Living Tissues. Springer-Verlag,
New York, pp. 211, 1981.
[81] Hjortdal, J.O. ―Regional elastic performance of the human cornea.‖ J. Biomechanics,
Vol. 29, No. 7, pp. 931-942, 1996.
[82] Shin, T.J., Vito, R.P. et al. ―The distribution of strain in the human cornea.‖ J.
Biomechanics, Vol. 30, No. 5, pp. 497-503, 1997.
[83] Erkamp R.Q., Wiggins P, Skovoroda, A.R. et al. ―Measuring the elastic modulus of
small tissue samples.‖ Ultrasound Imaging, 20:17–28, 1998.
[84] Chen E.J., Novakofski J., Jenkins W.K., et al. ―Young‘s modulus measurements of soft
tissues with application to elasticity imaging.‖ IEEE Trans. UFFC 43:191–4, 1996.
[85] Zeng, Y., Yang, J., et al. ―A comparison of biomechanical properties between human
and porcine cornea.‖ J. Biomechanics, Vol. 34, No. 5, pp. 533-537, 2001.
[86] Jue B, Maurice DM. The mechanical properties of rabbit and human cornea. J.
Biomech. 1986;19:847–853.
[87] Woo SLY, Kobayashi AS, Schlegel WA, Lawrence C. Non-linear Properties of intact
cornea and sclera. Exp. Eye Res. 1972; 14: 29-39.
[88] Nyquist GW. Rheology of the cornea: Experimental techniques and results.
Experimental Eye Research. 1968; 7: 183-188.
[89] Hjortdal JO. On the biomechanical properties of the cornea with particular reference to
refractive surgery. The Ophthalmological Journal of the Nordic Countries. 1998;
76(225): 1-23.
[90] Andreassen TT, Simonsen AH, Oxlund H. Biomechanical properties of keratoconus
and normal corneas. Exp. Eye Research. 1980; 31: 435-441.
[91] Liu, J., Roberts, C. J., 2005, ―Influence of corneal biomechanical properties on
intraocular pressure measurement: Quantitative analysis,‖ J. Cataract Refract Surgery,
31, pp. 146-155.
[92] Hoeltzel DA, Altman P, Buzard K, Choe K. Strip extensiometry for comparison of the
mechanical response of bovine, rabbit, and human corneas. J. Biomech. Eng.
1992;114(2):202–215.
[93] Daxer A, Misof K, Grabnerj B, Ettl A, Fratzl P. Collagen fibrils in the human corneal
stroma: Structure and aging. Investigative Ophthalmology and Visual Science,
1998;39(3):644-648.
[94] Malik NS, Moss SJ, Ahmed N, Furth AJ, Wall RS, Meek KM, Ageing of the human
corneal stroma: structural and biochemical changes, Biochimica et Biophysica Acta.
1992;1138:222-228.
[95] Wollensak G, Spoerl E, Seiler T. Stress-strain measurements of human and porcine

corneas after riboflavin–ultraviolet-A-induced cross-linking. Journal of Cataract and
Refractive Surgery. 2003;29:1780-1785.
[96] Anderson, K., Elsheikh, A. and Newson, T., Application of Structural Analysis to the
Mechanical Behaviour of the Cornea, Journal of Royal Society – Interface, 1, 2004, 1-
13
[97] Greene, P.R. ―Stress-strain behavior for curved exponential strips.‖ Bulletin of
Mathematical Biology, Vol. 47,No. 6, pp. 757-764, 1985.
[98] Elsheikh A, Anderson K. Comparative study of corneal strip extensometry and inflation
tests. Journal of Royal Society – Interface. 2005;2:177–185.
[99] Naor J, Slomovic AR, Chipman M, Rootman DS. A randomized, double-masked
clinical trials of Optisol-GS vs Chen medium for human corneal storage. Arch.
Ophthalmology. 2002;120:1280-1285.
[100] Arndt C, Reese S, Kostlin R. Preservation of canine and feline corneoscleral tissue in
Optisol GS. Veterinary Ophthalmology. 2001;4(3):175-182.
[101] Kaufman HE, Beuerman RW, Steinemann TL, Thompson HW, Varnell ED. Optisol
corneal storage medium. Arch. Ophthalmol. 1991;109(6):864-868.
[102] Wagoner MD, Gonnah E. Corneal graft survival after prolonged storage in Optisol-GS.
Cornea. 2005;24(8): 976-979.
[103] Orssengo GJ, Pye DC. Determination of the true intraocular pressure and modulus of
elasticity of the human cornea in vivo. Bull Math. Biol. 1999; 61:551-572.
[104] Elsheikh, A. and Alhasso, D., Mechanical Anisotropy of Porcine Cornea and
Correlation with Stromal Microstructure, Accepted for publication in Experimental Eye
Research.
[105] Bryant MR, Szerenyi K, Schmotzer H, McDonnell P. Corneal tensile strength in fully
healed radial keratotomy wounds. IOVS. 1994;35(7):3022-3031.
[106] Elsheikh A, Wang D. Numerical modelling of corneal biomechanical behaviour.
Computer Methods in Biomechanics and Biomedical Engineering. 2007;10(2):85-95.
[107] Fernández DC, Niazy AM, Kurtz RM, Djotyan GP, Juhasz T. A finite element model
for ultrafast laser-lamellar keratoplasty. Annals of Biomedical Engineering.
2006;34(1):169-183.
[108] Radner W, Zehetmayer M, Skorpik C, Mallinger R. Altered organization of collagen in
the apex of keratoconus corneas. Ophthalmic Research. 1998;30(5):327-332.
[109] Meek KM, Tuft SJ, Huang Y, Gill PS, Hayes S, Newton RH, Bron AJ. Changes in
collagen orientation and distribution in keratoconus corneas. IOVS. 2005;46:1948-1956.
[110] Luce DA. Determining in vivo biomechanical properties of the cornea with an ocular
response analyzer. Journal of Cataract Refractive Surgery. 2005;31:156-162.
[111] Shah S, Laiquzzaman M, Bhojwani R, Mantry S, Cunliffe I. Assessment of the
biomechanical properties of the cornea with the Ocular Response Analyzer in normal
and keratoconic eyes. IOVS. 2007;48:3026-3031.
[112] Zienkiewicz OC, Taylor RL. The finite element method. McGraw-Hill Book Company
Europe, Berkshire. 2000.
[113] Hamri O, Léon J-C, Giannini F, Falcidieno B. Exploiting features for finite element
model generation. Mathematics in Industry. 2006;8:585-589.
[114] Armstrong CG. Modelling requirements for finite-element analysis. Computer-aided
design. 1994;26(7):573-578.
110 Ahmed Elsheikh
[115] Babuška I, Strouboulis T. The finite element method and its reliability. Oxford Science
Publications. 2001.
[116] McPhee TJ, Bourne WM, Brubaker RF. Location of the stress-bearing layers of the
cornea. IOVS. 1985;26:869-872.
[117] Kotecha A, Elsheikh A, Roberts C, Zhu H, Garway-Heath DF. Corneal thickness- and
age-related biomechanical properties of the cornea measured with the Ocular Response
Analyzer. IOVS. 2006;47(12):5337-5347.
[118] Elsheikh, A., Alhasso, D., Kotecha, A. and Garway-Heath, D.F., Assessment of the
Ocular Response Analyzer as a tool for intraocular pressure measurement, Journal of
Biomechanical Engineering. In Press.
[119] Elsheikh, A., Wang, D., Kotecha, A., Brown, M. and Garway-Heath, D., Evaluation of
Goldmann applanation tonometry using a nonlinear finite element ocular model, Annals
of Biomedical Engineering, 34(10), 2006, 1628-1640.
[120] Pandolfi A, Manganiello F. A model for the human cornea: constitutive formulation
and numerical analysis. Biomechanics and modeling in mechanobiology.
2006;.5(4):237-246.
[121] Pinsky PM, van der Heide D, Chernyak D. Computational modeling of mechanical
anisotropy in the cornea and sclera. Journal of Cataract and Refractive Surgery.
2005;31:136-145.
[122] Curtin BJ, Physiopathologic aspects of scleral stress-strain. Trans. Am. Ophthalmol.
Soc. 1969;67:417-461.
[123] Philips JR, McBrien NA. Form deprivation myopia: elastic properties of sclera.
Ophthal. Physiol. Opt. 1995;15(5):357-362.
[124] Downs JC, Suh JKF, Thomas KA, Bellezza AJ, Hart RT, Burgoyne CF. Viscoelastic
material properties of the peripapillary sclera in normal and early-glaucoma monkey
eyes. IOVS. 2005;46:540-546.
[125] Downs JC, Suh JKF, Thomas KA, Bellezza AJ, Burgoyne CF, Hart RT. Viscoelastic
characterization of peripapillary sclera: material properties by quadrant in rabbit and
monkey eyes. Journal of Biomechanical Engineering. 2003;125:124-131.
[126] Barraglioli JL, Kamm RD. Measurements of the compressive properties of sclerol
tissue. IOVS. 1984;25:59-65.
[127] Spoerl E, Boehm AG, Pillunat LE. The influence of various substances on the
biomechanical behavior of lamina cribrosa and peripapillary sclera. IOVS.
2005;46:1286-1290.
[128] Norton TT, Rada JA. Reduced extracellular matrix in mammalian sclera with induced
myopia. Vision Research. 1995;35(9):1271-1281.
[129] Guggenheim JA, McBrien NA. Form-deprivation myopia induces activation of scleral
matrix metalloproteinase-2 in tree shrew. IOVS. 1996;37:1380-1395.
[130] McBrien NA, Gentle A. Role of the sclera in the development and pathological
complications of myopia. Progress in Retinal and Eye Research. 2003;22(3):307-338.
Chapter 3
BIOMECHANICS CONCEPTS OF BONE-ORAL

IMPLANT INTERFACE
Ahmed Ballo1,2 and Niko Moritz3

1
Department of Biomaterials, Institute of Clinical Sciences,
The Sahlgrenska Academy, Gothenburg University, Gothenburg, Sweden
2
Department of Prosthetic Dentistry and Biomaterials Science,
Institute of Dentistry, University of Turku, Turku, Finland
3
Orthopedic Research Unit, Department of Orthopedic Surgery and Traumatology,
University of Turku, Turku, Finland
ABSTRACT
Osseointegrated implants are actually replacements for natural teeth, and, like natural
teeth, they are exposed to various forces. The success of osseointegration is based on the
clinical outcome; clinicians must ensure that the stresses that the superstructure, implant,
and surrounding bone are subjected to are within the tolerable limits of the various
components. Structural compatibility is the optimum adaptation to the mechanical
behavior of the host tissues. Therefore, structure compatibility refers to the mechanical
properties of the implant material, such as elastic modulus, strength, implant design and
optimal load transmission (minimum interfacial strain mismatch) at the implant/tissue
interface, which is the key to the successful functioning of the implant device.
This chapter reviews some of the reaction, properties and characteristics of the bone
and explains how the bone-implant interface will react under loading condition. The
chapter also includes characteristics, properties and other important information about the
implant biomaterials and implant coating.
1. INTRODUCTION
Biomechanics is the application of engineering mechanics (statics, dynamics, strength of
materials and stress analysis) to the solution of biological problems (Fung, 1981).
112 Ahmed Ballo and Niko Moritz
Biomechanics pertains to clinical dentistry because the teeth and jaw perform biomechanical
activities during mastication.
An osseointegrated oral implant system consists of implants that connect a prosthesis to
the jaw bone. In all incidences of functional loading, forces are first introduced to the dental
prosthesis and then reach the bone-implant interface via the implant to distribute it over the
bone. The primary function of bone-implant interface is to provide a safe and effective load
transfer from the dental prosthesis to the bone. This process and the consequences of force
transmission into supporting bone depends on the nature of applied force (amplitude,
direction, rate and frequency), the design of implants (i.e., shape, surface macro- and micro-
architecture), the biological and biomechanical properties of the bone-implant interface and
the reaction of bone tissue to the mechanical environment created by loading of the implant.
Understanding the nature of the loading forces and the mechanism of such forces‘
transfer to the biological tissues have been suggested as factors in the development of implant
to tissue interface and, indeed, implant longevity (Lemons and Bidez, 1991).
Bone mechanical properties are basic parameters that reflect the structure and function of
bone. The mechanical behavior of bone in normal physiological situations is similar to that of
an elastic material with no visible change in external appearance. Also, bone has the ability to
repair itself by adaptive mechanisms. Although a dental implant is a dead object, for the
transmission of mechanical chewing forces into the bone the implant should interact
mechanically with biological tissue.
In order to do this successfully, there is a need to expand the knowledge related to the
biomechanics of bone and mechanical parameters of biomaterials that determine load on oral
implant units so the implant design will comply with the mechanical principles of the
biological tissue.
2. FORCES AND MOMENTS

Oral forces applied to human dentition are induced due to mastication and biting. Biting
is understood as the act of breaking food into large pieces that will be further chewed. In this
process, a tooth is subjected to what is called occlusal loads. In mastication, the forces are
shorter in duration and less intense.
Force is defined as a ―push or pull‖ that results from physical contact between two
objects. Forces acting on both teeth and oral implants are referred to as vector quantities; this
means that force quantities have direction as well as magnitude. Mastication mainly induces
vertical load in the dentition. The magnitude of masticatory forces applied to teeth varies
depending on the location in the jaws, with greater forces in the posterior regions. However,
transverse forces are created by horizontal motion of the mandible and the inclination of teeth
cusps.
The values of forces acting on teeth have been extensively addressed in the literature
(Anderson, 1953; Cimini et al., 2000; Duyck et al., 2000; Gibbs et al., 1981; Howell and
Brudevold, 1950; Neill et al., 1989; Waltimo and Kononen, 1993).
The magnitude of occlusal forces varies depends on tooth position in the mouth (Rangert
et al., 1995). The mean maximal occlusal forces in an incisal area have been reported to range
Biomechanics Concepts of Bone-Oral Implant Interface 113
from 264 N to 370 N (Laurell, 1985; Paphangkorakit and Osbom, 1997), whereas the average
maximum biting force in the posterior area can be around 800 N (Van Eijden, 1991).
During clinical loading of an oral implant, the direction of forces almost never coincides
along its central long axis, providing two main types of loading: (1) axial force and (2)
bending moment.
These two types of loading are completely different in nature. The axial force is more
favorable, as it distributes stress more homogenously throughout the implant, while the
bending moment exerts stress gradients in the implant as well as in the crestal bone around
dental implants which may act as a fulcrum point for lever action when a force is applied
(Smedberg et al., 1996; Richter, 1998), indicating that peri-implant tissue could be more
susceptible to crestal bone loss by applying forces.
This bending moment is the force times the orthogonal distance between the force
direction line and the counter-acting support (Figures 1 and 2). The longer the distance, the
greater the bending moment will be (White et al., 1994).
Rangert and his colleagues exemplified the nature of force versus moment by using the
―see-saw‖ model (Rangert et al., 1989). This model illustrates the beam balance of the
cantilever prosthesis under the loading; the heavier load should be applied closer to the
fulcrum than the lighter load applied on the opposite side.
The factor that determines this equilibrium of different forces is the ratio of the lever
arms. The product of force × lever arm must be equal at both sides of the fulcrum.
Figure 1. Different loading types from occlusion acting on implant prosthesis.

Figure 2. Forces within the implant compensating the bending moment from a vertical occlusal force
(bending moment on the implant).
The product of force and the distance to the force axis is called the bending moment, thus
Force × Lever arm = Bending moment
The optimum situation for the distribution of the bending moment along the implant is
placement of the implants as close together as possible. In Figure 2, the distal end of the
prosthesis thus acts as the fulcrum in this situation. The bending moment on the implant
therefore can induce extremely high internal force in the screw of the abutment, which may
easily overload the screw and might end up with screw fracture. Although the gold screw is
designed to be the weakest link in the implant-abutment-prosthesis connection, its breakage
may considerably complicate additional fixed prosthodontic treatment.
Also, the excessive bending moments may cause stress concentration and micro-fractures
in alveolar bone and even implant fractures.
3. BONE BIOMECHANICAL PROPERTIES

The bone structure can be exemplified by structure of the bridges. The visible structure
when looking to the distance view of the bridge (Figure 3) is the long beam, arches and
pillars. These elements represent the large scale level, while in close view the elements have a
hierarchy like the structure that reinforces the bridges.
Figure 3. Bridge design applying the principle of structural hierarchy for bridge reinforcement.
Bone tissue is a highly specialized form of connective tissue of mesenchymal origin with
a hierarchical structure (Kaplan et al., 1994 and Rubin et al., 2004). The structure is different
at different scales (Figure 4). Structure and microarchitecture are determinant aspects of bone
strength essential elements for the assessment of bone mechanical properties.
Figure 4. The hierarchical structure of cortical bone.
Bone is not homogenous; its physical properties are highly variable according to species,
age, sex, liquid content, type of bone (e.g., femoral, mandibular, cortical, cancellous) and
even according to the location within the bone from which the sample is taken (Katz, 1995).
Microarchitecture seems to be a determinant of bone fragility independent of bone

density, and important in order to understand the mechanisms of bone fragility (Dalle
Carbonare and Giannini, 2004). Among other factors, microarchitecture is affected by
loading. Loading also affects bone mass and size and it is therefore important for the
maintenance of bone strength during normal aging (Mosekilde, 2000).
Bone tissue is arranged in two macro-architectural forms: (I) trabecular (or cancellous or
spongy) and (II) cortical (or compact). These differences in macroarchitecture have been used
to derive a clinical classification of bone type in the dental implant field, based on the relative
proportion of cortical to trabecular bone (that is, where Class 1 bone is predominantly cortical
as in the anterior mandible, while Class 4 bone is almost all trabecular as found in the
posterior maxilla) (Leckholm and Zarb, 1985).
Cortical bone is an anisotropic material, meaning that its mechanical properties vary
according to the direction of load (Currey 1984) (Figure 5). Cortical bone is often considered
an orthotropic material. Orthotropic materials are a class of anisotropic materials
characterized by three different Young‘s moduli E1, E2, E3 according to the direction of load,
ν ν ν ν ν ν
three shear moduli G12, G13, G23 and six Poisson's ratios 12, 13, 23, 21, 31, 32.
Bone can be either woven (coarse-fiber bone) or lamellar (layered). Woven bone is weak,
with a small number of randomly oriented collagen fibers, but forms quickly and without a
pre-existing structure during periods of repair or growth. Lamellar bone is stronger, formed of
numerous stacked layers and filled with many collagen fibers parallel to other fibers in the
same layer.
Figure 5. Comparison between the mechanical behaviour of isotropic and anisotropic materials.
After a break, woven bone quickly forms and is gradually replaced by slow-growing
lamellar bone on pre-existing calcified hyaline cartilage through a process known as ―bony
substitution.‖
It has been observed that there is a relationship between bone structures and applied
loads. In 1892, Wolff found that the orientation of trabeculae coincides with the direction of
the stress trajectories. He proposed that bone loading is somehow sensed and that the bone
adapts its structure accordingly. This principle of functional adaptation is generally known as
―Wolff‘s Law‖ (Wolff, 1892). It is occurs in conditions of disuse when bone is lost (Bauman
et al., 1999; Vico et al., 2000; Zerwekh et al., 1998), and in overloading, which causes a gain
in bone mass (Suominen, 1993). It also occurs after fracture healing (Wolff, 1892) and in
relation with implant incorporation (Guldberg et al., 1997).
The bone acts as if it has some sensors that can measure the internal load and activate the
bone cells to carry out the bone remodeling. They are based on the separate actions of bone
resorbing cells called osteoclasts and bone forming cells called osteoblasts. Bone remodeling
(Figure 6), appears to be governed by a feedback system in which the bone cells sense the
state of strain in the bone matrix around them and either add or remove bone as needed to
maintain the strain within normal limits. Osteocytes may play an important role here.
Several studies revealed that these cells respond to mechanical stimulation (Skerry et al.,
1989; Klein-Nulend et al., 1995). Together with the lining cells they form a system that seems
well equipped for signal transduction (Cowin et al., 1991). It could be that mechanically
induced osteocyte signals are transferred through the canaliculi to the bone surface where
they control osteoclast and osteoblast activity (Burger and Klein-Nulend, 1999).
Figure 6. Schematic diagram illustration bone remodeling process.
Bone fractures can be produced by either a single load or the repeated application of load.
A fracture will result from the application of a single load if that load produces a stress or
strain that exceeds the ultimate strength of the bone. The repeated application of a lower load
may also produce a fracture, such a fracture is called a fatigue fracture and is typically
produced by either lower repetition of high loads or high repetition of relatively low loads.
4. PRINCIPLES OF FORCES TRANSFER TO BONE-IMPLANT INTERFACE

The forces generated during mastication influence the teeth, as well as the periodontium,
the jaw bones and the temporomandibular joints. Unlike nature teeth, osseointegrated
implants are ankylosed to surrounding bone without the periodontal ligament (Figure 7),
which provides mechanoreceptors as well as shock-absorbing function (Schulte, 1995).
Successful implant osseointegration and its clinical longevity depend upon the way
mechanical stresses are transferred to the surrounding bone. This force transfer from the
implant to the surrounding bone is influenced by the type of loading that occurs (i.e.,
intermittent, continuous), the bone-implant interface (i.e., direct contact or a gap interface),
the length and diameter of the implant, the implant shape, the surface texture of the implant
and the quality and quantity of the surrounding bone (Bechtold et al., 2001; Jones et al., 2001;
Mouzin et al., 2001; Skripitz and Aspenberg, 2001; Toksvig-Larsen et al., 2001).
Figure 7. Schematic diagram of an endosseous dental implant and a tooth. Note that the implant lacks
the periodontal ligament and is dependent on direct bone support.
The distributions of forces and moments throughout the structure depend on the relative
flexural rigidities of the components and on the rigidity of the connections. Some
assumptions about rigidity must be made in order to simplify the system but still retain the
features necessary to give clinically useful results.
The routine loading protocol established by Branemark and his colleagues was
empirically based and used three months for the mandible and six months for the maxilla
(Branemark et al., 1985). The primary aim of such an approach is to create a healing
environment at the bone–implant interface that is capable of nurturing regeneration and
osseointegration rather than repair and fibrous encapsulation. However, clinical and
experimental studies demonstrate that osseointegration can be achieved when healing under
load is allowed.
The bone tissue response to immediately loaded oral implants has been described by
various investigators (Hashimoto et al., 1988; Lum et al. 1991; Akagawa et al., 1993; Piattelli
et al., 1993; Sagara et al., 1993; Piattelli et al., 1997; Corso et al., 1999) and Szmukler-
Moncler and his colleagues have given a literature review concerning the timing of loading
and effect of micromotion on bone-implant interface (Szmukler-Moncler et al., 1998).
Clinical (Quirynen et al., 1992; Naert et al., 1992; Baron et al., 2005) as well as
experimental studies (Miyata et al., 2000, 2002) indicate that marginal bone loss at the oral
implant may be associated with high occlusal stress on implants. In poor bone conditions, the
mismatch of stiffness between bone and metallic implant may lead to implant failure
(Lemons, 1998).
This occurs when the tensile or compressive load exceeds the physiological limit of bone
tolerance and causes microfracture at the bone-to-implant interface, or initiates bone
resorption (Brunski, 1999).
While overloading may be manifested by the application of repeated single loads, which
causes micro-fractures within bone tissue, continuous application of low loads may also lead
to failure, namely, fatigue fracture. However, stress-shielding (underload) at the interface
have also been suggested as a contributing factor for bone resorption. Therefore oral implant
should be designed so that it induces a mechanical stimulation in the surrounding bone and
high stress peaks do not arise in the bone.
On small scale this micromotion at the implant-bone interface can have two principle
effects on cellular and extracellular components of bone. First, the micromotion can lead to a
disruption of the bone-cell/implant contacts and therefore has the ability to disturb the cell
reaction by a detachment; or second the micromotion can lead to a deformation of osteoblast
cells attached to the surface in a strain-related manner.
The bone tissue formation and mineralization by osteoblast cells are dependent on the
local mechanical interaction at the cell-implant surface zone is generally accepted (Szmukler-
Moncler et al., 1998). Cells generate, transmit and sense mechanical tension, they use these
forces to control their shape and behavior. Cells generate tensile forces within contractile
microfilament in cytoskeleton by use of an actomyosin filament sliding mechanism (Chirucel
et al. 1998).
The cells extend lamellopodia (Figure 8) that adhere to the implant surface formed the
sites act as mechanosensory ―devices,‖ where internal contractile forces or externally applied
force can regulate the assembly of the adhesion site (Geiger and Bershadsky, 2001).
Figure 8. Cell migration. The cells extend lamellopodia that adhere to the implant surface.
The cytoskeleton will subject the adhesion site to a tensile force. The force developed
depends on the resilience of the implant material, the stiffer the material, the greater the force.
The mechanical strength of both the adhesion site and the part of cytoskeleton under tension
are adapted to the force, a consequence of this adaptation of force and strength to the
resistance met is that cells tend to migrate in the direction of increasing material stiffness (Lo
et al., 2000, Dischler et al., 2005).
Carter and Giori suggested that proliferation and differention of the osteoblast cells
responsible for peri-implant tissue formation are regulated by the local mechanical
environment according to the tissue differentiation hypothesis proposed by Frost and his
colleagues for callus formation (Carter and Giori, 1991).
Recently development in implant surfaces has shown improvement in the mechanical
interaction with the bone. Figure 9 illustrates the different scale levels of bone-implant
interaction all the way from the overall design down to the molecular level.
Figure 9. Mechanical interaction between implant and bone on five different scale levels (from the
design level to the molecular level).
5. BULK MECHANICAL PROPERTIES OF IMPLANT BIOMATERIALS

The biocompatibility is a descriptive term which indicates the ability of the material to
perform with an appropriate host response, in a specific application (Black and Hasting,
1998). This definition has been extended and distinguished between the surface and structural
compatibility of an implant (Wintermantel andMayer, 1995). Surface compatibility meaning
the chemical, biological, and physical (including surface morphology) suitability of an
implant surface to the host tissues. Structural compatibility is the optimum adaptation to the
mechanical behavior of the host tissues. Therefore, structure compatibility refers to the
mechanical properties of the implant material, such as elastic modulus, strength, implant
design and optimal load transmission (minimum interfacial strain mismatch) at the
implant/tissue interface (Hedia, 2002).
Any implant must be constructed from a biomaterial. The biological performance of the
chosen biomaterial will be of concern. Biomaterials are generally composed of a wide
spectrum of materials. These implant biomaterials can be broadly classified into metallic,
ceramic, polymeric, and composite materials.
An improperly chosen material can lead not only to failure of the part or a structure but
also to unnecessary cost. Mechanical characterization is the classification of materials
according to their mechanical properties. Table 1 illustrates some mechanical properties of
these materials (O‘Brien and William, 2002; Bouillaguet et al., 2006).
Table 1. Comparison of mechanical properties of different implant materials
Property Titanium Alumina Zirconia Hydroxyapatite FRC*

/Materials
Strength 800–1000 400–600 900–1200 115–150 700–1000

(MPa)
Young’s 110–120 380 210 85 20–40

modulus (GPa)
Hardness (HV) 100–140 2200 1200–1500 300–400 70–80

*
Unidirectional fiber-reinforced.
The main considerations in selecting metals and alloys for biomedical applications are
biocompatibility, appropriate mechanical properties, and corrosion resistance. The high
tensile and fatigue strength of metals, compared with ceramics and polymers, make them the
materials of choice for implants that carry mechanical loads. However, implants are made of
metals which are stiffer than bone which they contact. This leads to large percent of the load
normally transmitted by the bone being borne by the prosthesis instead. This shielding can
lead to bone disuse atrophy (Chrisman and Snook, 1968). Using an implant material which is
biomechanically more suitable, which it is strong enough but also flexible so that it can
reduce the interfacial stresses that causes fatigue fracture or resorption of the bone.
The flexibility or stiffness of a material may be described by the so-called modulus of
elasticity (Young‘s modulus). The elastic modulus is the constant that relates the stress and
the strain in the linear elastic region where elastic deformation of a material occurs (Figure
10) (Van Noort, 2002).
Commercial-purity titanium (cp Ti) comes in a number of grades, with the mechanical
properties being somewhat different for each grade. Small changes in the oxygen and iron
contents greatly influence the mechanical properties, for example, Grade 3 cp Ti has Fe and O
contents (maximum) of 0.3 and 0.35 %, respectively, and a yield strength of 380 MPa, while
Grade 4 has Fe and O contents 0.5 and 0.4%, respectively, and a yield strength of 483 MPa.
The Ti-6%Al-4%V alloy, however, has a much higher value for yield strength (795 MPa), so
this alloy is often chosen for dental implant cases where high strength is necessary (Mishra et
al., 1996). Compared with the elastic moduli of either stainless steel or cobalt-chromium, Ti
and its alloys have much lower moduli that are still almost an order of magnitude higher than
that of bone.
Figure 10. Average stress-strain curve for different materials.
Ceramic are materials composed of metallic and nonmetallic elements held together by
ionic and/or covalent bonds. Ceramic material has already been utilized as dental implant
material like aluminum oxide (Al2O3) (Schulte, 1984). This material osseointegrated well but
unfortunately, the biomechanical properties of the implants were not sufficient for long-term
load. As a result, this material was withdrawn from the market as dental implant material.
Recently, Zirconia has been introduced as a new ceramic dental implant material. As a metal
substitute, Zirconia possesses good chemical and physical properties, like low corrosion
potential, low thermal conductivity (Drouin and Cales, 1994; Piconi et al., 1998; Richter et
al., 1994). Furthermore, its biocompatibility and biomechanical properties as dental implant
material has been extensively investigated (Albrektsson et al., 1985; Ichikawa et al., 1992;
Kohal et al., 2006). However, these implant materials have very high elastic moduli compared
to that of human bone (17–24 GPa).
Hydroxyapatite (HA) Ceramic have been investigated extensively and used for dental
implant applications for the past 30 years. HA Properties depend on its porosity. Clinical
studies demonstrated that HA ceramics still remain the most biocompatible bone implant
material known and possess the added feature of becoming strongly bonded to living bone
through natural-appearing bonding mechanisms (Chang et al., 1996), yet the range of
problems included fractures during surgery, fractures after loading (Jarcho, 1992). Therefore,
HA ceramics are not ideal materials for permanent implant devices. However, bioactive
ceramic coatings on metal implants have kept ceramics as a key component in dental
implantology (Krauser et al., 1990; Ogiso et al., 1996). However, delamination of the ceramic
layer from the metal surface can create serious problems and lead to implant failure.
Polymers are widely used materials in biomedical applications. Polymers are used in drug
delivery systems and as a scaffolding material for tissue engineering applications. The
polymeric materials have molecular structures completely different from living substances
which make them generally more stable in the tissues.
Compared with metal and ceramics, polymers have much lower strength and moduli but
they can be deformed to a greater extent before failure. Consequently, polymers are generally
not used in biomedical applications that bear loads. Ultra-high-molecular-weight polyethylene
is an exception, as it is used as a bearing surface in hip and knee replacements.
In high-load trauma applications such as bone fracture plates, screws and intramedullary
nails, the strength of Polyetheretherketone (PEEK) polymer allows it to be a true replacement
of metals (Kurtz and Devine, 2007). But polymeric are not commonly seen for dental
implants yet.
In 1969, Hodosh placed custom-made polymer implant directly into the extraction socket
of tooth for the first time (Brown et al., 1969; Ehrlich and Azaz, 1975; Hodosh et al., 1969;
Hodosh et al. 1973). Since then most studies have been done with experimental animals but
some work has been published with selected human patients as well (Hodosh et al., 1969). It
has been reported that attachment of polymer implant can be achieved with connective tissue
capsule resembling the periodontal ligament (Brown et al., 1969). However, due to the high
failure rate of 60% after the seven years follow up of 10 patients, this acrylic resin implant
was not recommended for clinical use (Hodosh et al., 1970). At the same time, Brånemark
introduced good results on the osseointegrated titanium implants (Brånemark et al., 1977).
Recently, fiber-reinforced composite (FRC) has been evaluated as a new material for oral
and maxillofacial implants (Ballo et al., 2007; Ballo et al., 2008a and 2008b; Ballo et al.
2009). The pre-clinical studies concerning the mechanical properties of biostable glass FRC
implants have shown good performance in the laboratory environment (Ballo et al., 2007 and
2008a). Both in vitro and in vivo studies have shown that FRC material is biocompatible in
the cell culture environment and preosteoblasts have been shown to mature and proliferate
well on the FRC substrates (Ballo et al., 2008b). The rigid bone-implant interface has been
observed in a histological investigation without fibrous connective tissue between the bone
and the FRC implant (Ballo et al., 2009). These studies used well-polymerized FRC material
in combination with bioactive glass.
Although these are promising results with FRC in pre-clinical studies, many questions
remain to be answered before FRC implants can be introduced into clinical use. In the future,
further studies are needed to evaluate the bone remodeling process of the FRC implants under
loading condition.
8. IMPLANT SURFACE COATING

Various surface coating s have been extensively evaluated in implant dentistry. Implant
coatings have strong impact factor on the strength of bone-implant interface. Chemical
bonding is descripting when bioactive ceramics chemically bond to the surrounding bone
(Nakamura et al., 1985).
Bioactive ceramic coatings on metal implants have kept ceramics as a key component in
dental implantology (Krauser et al., 1990; Ogiso et al., 1996).
Until the introduction of the concept of bioactive materials, surface reactivity was
considered an unwanted property for an implant material. However, some surface reactive
(biologically active, bioactive) materials have shown the ability to form an interfacial
chemical bond with surrounding tissues through a series of biophysical and biochemical
reactions, causing ―bioactive fixation‖ of the implant (Cao and Hench 1996). Bioactive
materials can be biostable (i.e., synthetic hydroxylapatite) or bioresorbable (i.e., bioactive
glasses and glass-ceramics). Some bioactive ceramics like bioactive glasses of certain
compositions have been claimed to have a real chemical bonding ability with soft tissues
(Wilson, 1981 and 1990). There are indications that titanium oxides are also capable of
forming an adherent bond with soft tissues (Areva, 2004; Rossi, 2007; Paldan, 2008).
However, bioactive ceramics in the bulk form are not suitable for load-bearing
applications as their flexural strength, strain-to-failure and fracture toughness are less than
that of bone and their elastic moduli are greater than that of bone. For these reasons they are
usually applied as coatings on metallic implants that possess superior mechanical properties.
Although in the past 30–40 years many studies have been carried out in the field of
biomaterials to achieve better tissue attachment, today the majority of implants are still made
of biologically inactive materials. This is mainly due to the fact that either these materials
have a history of reliable performance or simply because nothing better is available. For
example, titanium and its alloys are widely used in clinical practice. Titanium is a highly
reactive metal and except for certain artificial conditions is covered by a thin (2–10 nm)
surface oxide layer. It is the chemical stability of this oxide layer (mainly TiO2) that provides
the superior corrosion resistance of the implants made of titanium and its alloys. Although
Young‘s modulus of titanium alloys (~100 GPa) differs considerably from that of bone (7–25
GPa for cortical bone, 0.05 GPa–0.5 GPa for cancellous bone) (Hench 1994), these materials
are well integrated into bone through morphological or biological fixation and have sufficient
strength for load-bearing applications. However absence of a physico-chemical bond between
the implant and the host bone especially in the poor bone quality conditions poses a risk of
implant failure. Further, in the case of percutaneous or permucosal implants, such as dental
implants, lack of bonding to connective tissue and epithelium may help the pathogenic
microorganisms to migrate along the surface of the implant leading to infection. Bioactive
fixation between an implant and soft tissues could also prevent this (Jansen, 1992).
8.1. Calcium-Phosphate Coatings
To improve attachment between the surface of a medical implant and the surrounding
tissues, surface-reactive (bioactive) calcium-phosphate-based coatings can be applied to the
surface of an implant by different methods (Leeuwenburgh, 2008; Wolke, 1992;

Kangasniemi, 1994; Lacefield, 1988 and 1994,). Calcium-phosphates include a large group of
structurally related apatites (tricalcium phosphates, hydroxyapatite, fluorapatite, etc). At
present, coating of a medical implant surface with calcium-phospahte by plasma spraying is
the only method approved by the FDA (Food and Drug Administration) (Campbell, 2003).
8.2. Bioactive Glass Coatings
Originally introduced by Hench (Hench, 1971), silica-based bioactive glass materials are
slowly resorbing synthetic osteoconductive materials which are able to form strong chemical
bond with bone. Several methods have been applied to obtain a bioactive glass coating on the
load bearing part of the implants (Hench, 1993). Numerous attempts have been made to create
bioactive glass coatings on alumina (Greenspan and Hench, 1976; Griss et al., 1976), stainless
steel (Piotrowski et al., 1975; Schepers et al., 1989), Co-Cr-Mo alloy (Ducheyne et al., 1984;
Lacefield and Hench, 1986; Pajamäki et al., 1995) and titanium alloys (West et al., 1990;
Kitsugi et al., 1996). Application of a double glass coating has been suggested to solve the
problem of differences in thermal expansion coefficients.
Recently, several attempts were made to prepare bioactive glass coatings on titanium and
Ti6Al4V (Bloyer et al., 1999; Saiz et al., 2002). Application of a ground layer prepared from
inert glass with a thermal expansion coefficient close to that of Ti6Al4V provided good
adhesion to the substrate (Oku, 2001). This ground coating can be used in combination with
more surface reactive glass coatings (Gomez-Vega, 2000a), with embedded hydroxyapatite
and/or bioactive glass particles (Gomez-Vega, 2000b), or a sol-gel-derived silica coating
(Gomez-Vega, 2001). Reactive plasma spraying (Schrooten, 2000) or processing with infra-
red laser (Moritz, 2004) have also been attempted to create bioactive glass coatings on
titanium and its alloys.
9. BIOMECHANICAL CONSIDERATIONS FOR

COATED IMPLANT DESIGN
With respect to orthopaedic hip implants, a lot of knowledge has been accumulated
leading to identification of a number of important implant failure scenarios (Huiskes, 1993).
Although originally discussed with respect to orthopaedic implants, similar issues are valid
for other types of load-bearing medical implants, including dental.
Under repeated dynamic loading there is an accumulation of mechanical damage in
materials and materials‘ interfaces. When the interface stresses exceed the strengths of the
materials, the disruption of the implant is initiated by the growth of cracks at the interface.
This is followed by interface micromotion, bone resorption, fibrous encapsulation and
ultimate failure of the implant (Huiskes, 1993). For uncoated implants, failed bonding is
another important issue. If bone ingrowth or osseointegration does not occur, the resulting
motion of the interface in excess of 150 µm will always lead to the formation of the fibrous
membrane (Huiskes, 1993). The ―stress-shielding‖ phenomenon is often addressed in the
literature. After the implantation, it is the bone and the implant together that are engaged in
sharing of the load that is normally carried by bone alone. As the implant takes a fraction of
the load away from the bone, subnormal loading of the bone according to Wolff‘s law will
lead to its resorption in the form of osteopenia and cortical thinning. The mismatch of the
mechanical properties often leads to physiological underloading of the bone, thus increasing
the likelihood of adverse bone remodelling. However, it should be noted that, regardless of
the mechanical properties of the implant, just the addition of an implant would always cause
some degree of stress-shielding. Thus, it is impossible to completely eliminate stress-
shielding. Also, stress bypass and destructive wear are examples of other possible failure
scenarios (Huiskes, 1993). Moreover, all of these failure scenarios are interlinked, thus there
is a need for a systematic approach to implant design so that the entire scope of conflicting
requirements are considered. For example, using less-stiff implant materials will obviously
decrease the effect of stress-shielding but respectively increase the interfacial stresses, thus
promoting accumulated damage and failed ingrowth (Huiskes, 1993).
Regardless of all the possible benefits of bioactive fixation, addition of a bioactive
coating will be practical only if the coating is adequately strong to withstand the stresses the
implant is subjected to. This applies to implant-coating interface, coating-tissue interface and
the coating material itself. Moreover, in real clinical applications, the integrity of the
interfaces should be maintained over a long period of time.
Strong bonds between bioactive materials and periimplant bone is a well-known
phenomenon reported by many authors (Nakamura et al., 1985; Li et al., 1995; Andersson et
al., 1992). In a number of studies by Soballe, performance of hydroxyapatite coatings was
tested in different in vivo conditions, including gap-healing, osteopenic bone bed, bone
grafting and dynamic loading (Soballe, 1993), demonstrating adequate performance of the
coatings. In some studies (Cook, 1992a), selection of torsional loading as a biomechanical
testing method is based on the assumption that the torsional loads may influence the longevity
of dental implants. Moreover, as it has been demonstrated, in torsional loading the HA-coated
cylindrical implants failed mainly at the coating-substrate interface (Cook, 1992a). It has been
demonstrated that when load-bearing is not taken into consideration, the implantation of
transcortical implants in a femur bone does accurately simulate biological responses to the
implants placed in the mandible (Cook, 1992b). Thus, the results were expected to be more
clinically relevant than if push-out or pull-out testing was used. The later methods would have
also required conically- rather than cylindrically-shaped implants (Andersson, 1992), as
misalignment of an implant during testing might easily occur and lead to incorrect results.
Further, in the case of macro-textured cylindrical implants such as bioactive glass-coated
implants, pull-out or push-out tests were suggested to be unacceptable methods (Dhert and
Jansen, 2000).
A large number of testing methods exist to quantify the adhesion strength of the coating
to the substrate (Rickerby, 1988; Berndt and Lin, 1993). A number of the testing methods
(Manley et al., 1987; Ducheyne and Martens, 1988) are suggested by the FDA alongside
standardized testing methods. According to FDA guidelines for testing of the plasma-sprayed
coating (FDA 2000), the shear fatigue strength of surface coating should be tested out to at
least 107 cycles. The static shear strength of the surface/substrate interface should be tested
and the adhesion shear strength should exceed 20 Mpa. The static tensile strength of the
surface/substrate interface and the coating should exceed 22 Mpa in tensile strength. While
these tests are useful for validation of the adhesion quality of a coating, they are not
representative of loading conditions of a real implant. Thus, to understand the fracture
behavior of a coating under loading conditions, a combination of realistic mechanical testing,

acoustic emission measurements and finite element analysis (FEA) were suggested
(Schrooten, 1999 and 2000). This allows for qualitative and quantitative evaluation of
coatings on a medical implant. The pattern of load transfer depends on several mechanical
factors, such implant geometry, properties of the implant material, interfacial strength and
loading conditions; thus, finite element analysis has been found to be a valuable tool in
development of human joint implants (Huiskes, 1993; Prendergast, 1997). For dental
implants, FEA has also been extensively used (van Staden et al., 2006). FEA allows for
simulation of real biomechanical testing conditions to study stress distributions at the
interfaces of interest. However, FEA of coated implants has received less attention (Mihalko
et al., 1992; Evans et al., 1994; Evans and Gregson, 1994; Schrooten 1999 and 2000; Aoki et
al., 2006). Introduction of a low modulus interlayer between the implant and the surrounding
bone can totally change the pattern of load transfer. When this is done in a controlled way, the
performance of the implant can even be improved; however, other aspects of implant design
should also be taken into account (Evans and Gregson, 1994). Moreover, a recent FEA study
has indicated that when applied to less stiff materials such as fiber-reinforced composites, a
bioactive surface component will promote more stress-shielding than an uncoated implant
(Zhao et al., 2009).
REFERENCES
Akagawa, Y., Ichikawa, Y., Nikai, H. and Tsuru, H. (1993). Interface histology of early
loaded partially stabilized zirconia endosseous implant in initial bone healing. Journal of
Prosthetic Dentistry, 69, 599–604.
Albrektsson, T., Hansson, H.A. and Ivarsson, B. (1985). Interface analysis of titanium and
zirconium bone implants. Biomaterials, 6, 97–101.
Anderson, D.J. (1953). A method of recording masticatory loads. Journal of Dental Research,
32, 785-789.
Andersson, Ö.H., Liu, G., Kangasniemi, K. and Juhanoja, J. (1992). Evaluation of the
acceptance of glass in bone. Journal of Materials Science Materials in Medicine, 3, 145-
150.
Aoki, H., Ozeki, K., Ohtani, Y., Fukui, Y. and Asaoka, T., (2006). Effect of a thin HA
coating on the stress/strain distribution in bone around dental implants using three-
dimensional finite element analysis. Bio-Medical Materials and Engineering, 16, 157-
169.
Areva, S., Paldan, H., Peltola, T., Närhi, T., Jokinen, M. and Linden, M. (2004). Use of sol-
gel derived titania coating for direct soft tissue attachment, Journal of Biomedical
Materials Research Part A, 70, 169-178.
Areva, S., Peltola, T., Säilynoja, K., Laajalehto, K., Linden, M. and Rosenholm, J.B. (2002).
Effect of albumin and fibrinogen on calcium phosphate formation on sol-gel derived
titania coatings in vitro. Chemistry of Materials, 14, 1614-1621.
Ballo, A.M, Lassila, L.V., Vallittu, P.K. and Närhi, T.O. (2007). Load bearing capacity of
bone anchored fiber-reinforced composite device. Journal of Materials Science Materials
in Medicine, 18, 2025-2031.
Ballo, A.M., Lassila, L.V.J., Närhi, T.O. and Vallittu, P.K. (2008a). In vitro mechanical
testing of glass fiber-reinforced composite as dental implants. Journal of Contemporary
Dental Practice, 9, 41-48.
Ballo, A.M., Kokkari, A.K., Meretoja, V.V., Lassila, L.V., Vallittu, P.K. and Närhi, T.O.
(2008b). Osteoblast proliferation and maturation on bioactive fiber-reinforced composite
surface. Journal of Materials Science Materials in Medicine, 19, 3169-3177.
Ballo, A.M., Akca, E., Ozen, T., Lassila, L., Vallittu, P.K. and Närhi, T.O. (2009). Bone
tissue responses to glass fiber-reinforced composite implants—a histomorphometric
study. Clinical Oral Implants Research, 20, 608-615 .
Baron, M., Haas, R., Baron, W. and Mailath-Pokorny, G. (2005). Peri-implant bone loss as a
function of tooth-implant distance. International Journal of Prosthodontics, 18, 427-433.
Bauman, W.A., Spungen, A.M., Wang, J., Pierson, R.N. Jr and Schwartz, E. (1999).
Continuous loss of bone during chronic immobilization: a monozygotic twin study,
Osteoporosis International, 10, 123-127.
Bechtold, J.E., Mouzin, O., Kidder, L. and Soballe, K. (2001). A controlled experimental
model of revision implants: Part II. Implementation with loaded titanium implants and
bone graft. Acta Orthopaedica Scandinavica, 72, 650-656.
Berndt, C.C. and Lin, C.K. (1993). Measurement of adhesion for thermally sprayed materials.
Journal of Adhesion Science and Technology, 7, 1235-1264.
Black, J. and Hasting, G.W. (1998). Handbook of biomaterials properties. London, UK:
Chapman and Hall.
Bloyer, D.R., Gomez-Vega, J.M., Saiz, E., Mcnaney, J.M., Cannon, R.M., Tomsia, A.P.
(1999). Fabrication and characterization of a bioactive glass coating on titanium implant
alloys, Acta materialia, 47, 4221-4224.
Bouillaguet, S., Schutt, A., Alander, P., Schwaller, P., Buerki, G., Michler, J., Cattani-
Lorente, M., Vallittu, P.K. and Krejci, I. (2006). Hydrothermal and mechanical stresses
degrade fiber-matrix interfacial bond strength in dental fiber-reinforced composites.
Journal of Biomedical Materials Research Part B Appl Biomaterials, 76, 98-105.
Brånemark, P.I., Hansson, B.O., Adell, R., Breine, U., Lindström, J., Hallén, O., Ohman, A.
(1977). Osseointegrated implants in the treatment of the edentulous jaw. Experience from
a 10 year period. Scandinavian Journal of Plastic and Reconstructive Surgery and Hand
Surgery, 16, 130-136.
Brånemark, P.I., Zarb, G.A. and Alberktsson, T. (1985). Tissue integrated prostheses:
Osseointegration in Clinical Dentistry. Chicago: Quintessence Publishing Company.
Brown, B.H., Neff, P.A. and Tylenda, A. (1969). Acrylic resin tooth implants in rhesus
monkeys: a preliminary report. Journal of Prosthetic Dentistry, 22, 367-370.
Brunski, J.B. (1999). In vivo bone response to biomechanical loading at the bone/dental-
implant interface. Advances in Dental Research,13, 99-119.
Burger, E.H. and Klein-Nulend, J. (1999). Mechanosensory transduction in bone – role of the
lacuno-canalicular network. The Journal of the Federation of American Societies for
Experimental Biology, 13, 101-112.
Campbell, A.A. (2003). Bioceramics for implant coatings, Materials Today, 11, 26-30.
Cao, W. and Hench, L.L. (1996). Bioactive materials, Ceramics International, 22, 493-507.
Carter, D.R., Giori, N.J. (1991). Effect of mechanical stress on tissue differentiation in the
bony implant bed. In: J.E. Davies (Ed.), The bone–biomaterial interface (pp. 367-376).
Toronto: University of Toronto Press.
Chang, Y.S., Oka, M., Nakamura, T., Gu, H.O. (1996). Bone remodeling around implanted
ceramics. Journal of Biomedical Materials Research, 30, 117-24.
Chicurel, M.E., Chen, C.S. and Ingber, D.E. (1998). Cellular control lies in the balance of
forces. Current Opinion in Cell Biology, 10, 232-239.
Chrisman, O.D. and Snook, G.A. (1968). The problem of refracture of the tibia. Clinical
Orthopaedics and Related Research, 60, 217-219.
Cimini, J.r, Gouvêa, P.H.P, Las Casas, E.B. and Cornacchia, T.P.M. (2000). Loads in teeth: a
critical review. In: Proceedings of the METMBS‘2000: The 2000 International
Conference on Mathematics and Engineering Techniques in Medicine and Biological
Sciences (2, pp. 649-655). Las Vegas, NV: CSREA Press.
Cook, S.D., Baffes, G.C., Palafox, A.J., Wolfe, M.W. and Burgess, A. (1992). Torsional
stability of HA-coated and grit-blasted titanium dental implants. Journal of oral
implantology, 18, 354-358.
Cook, S.D., Baffes, G.C., Wolfe, M.W. and Palafox, A.J. (1992). A comparison of femoral
and mandibular animal models for the evaluation of HA-coated implants. Journal of oral
implantology, 18, 359-365.
Corso, M., Sirota, C., Fiorelline, J., Rasool, F., Szmukler-Moncler, S. and Weber H-P (1999).
Evaluation of the osseointegration of early loaded free-standing dental implants with
various coatings in the dog model: periostest and radiographic results. Journal of
Prosthetic Dentistry, 82, 428–435.
Cowin, S.C., Moss-Salentijn, L. and Moss, M.L. (1991). Candidates for the mechanosensory
system in bone. Journal of Biomechanical Engineering, 113, 191-197.
Currey, J. (1984). The Mechanical Adaptations of Bone. New Jersey: Princeton University
Press.
Dalle Carbonare, L. and Giannini, S. (2004). Bone microarchitecture as an important
determination of bone strength. Journal of Endocrinoly Investigations, 27, 99–105.
Dhert, W.J.A., Jansen, J.A. (2000). The validity of a single pushout test. In Y. An, R.
Draughn (Eds.), Mechanical testing of bone at the bone-implant interface (PP. 477-488).
Boca Raton, FL: CRC Press.
Drouin, J.M. and Cales, B. (1994). Yttria-stabilized zirconia for improved hip joint head. In:
Ö.H. Andersson, A. Yli-Urpo (Eds.), Bioceramics 7 (pp. 387-394). London, UK:
Butterworth-Heinemann Publishers.
Ducheyne, P., Martens, M. and Burssens, A. (1984). Materials, clinical and morphological
evaluation of custom-made bioactive-glass-coated canine hip prostheses, Journal of
Biomedical Materials Research, 18, 1017-1030.
Ducheyne, P. and Martens, M. (1988). Orderly oriented wire meshes as porous coatings on
orthopaedic implants II: the pore size, interfacial bonding and microstructure after
pressure sintering of titanium (OOWM). Clinical Materials, 1, 91-98.
Duyck, J., Van Oosterwyck, H., De Cooman, M., Puers, R., Vander Sloten, J. and Naert, I.
(2000). Three-dimensional force measurements on oral implants: a methodological study.
Journal of Oral Rehabilitation, 27, 744-753.
Ehrlich, J. and Azaz, B. (1975). Immediate implantation of acrylic resin teeth into human
tooth sockets. Journal of Prosthetic Dentistry, 33, 205-209.
Evans, S.L., Lawes, K.R. and Gregson, P.J. (1994). Layered, adhesively bonded
hydroxyapatite coatings for orthopaedic implants. Journal of Materials Science Materials
in Medicine, 5, 495-499.
Evans, S.L., Gregson, P.J. (1994). Numerical optimization of the design of a coated,
cementless hip prosthesis. Journal of Materials Science Materials in Medicine, 5, 507-
510.
FDA (2000) Guidance for industry on the testing of metallic plasma sprayed coatings on
orthopedic implants to support reconsideration of postmarket surveillance requirements.
U.S. Department of Health and Human Services, Food and Drug Administration Center
for Devices and Radiological Health.
Fung, Y.C. (1981) Biomechanics Mechanical Properties of Living Tissues. New York:
Springer-Verlag.
Geiger, B. and Bershadsky, A. (2001). Assembly and mechanosensory function of focal
contacts. Current Opinion in Cell Biology, 13, 584-592.
Gibbs, C.H., Mahan, P.E., Lundeen, H.C., Brehnan, K., Walsh, E.K. and Holbrook, W.B.
(1981). Occlusal forces during chewing and swallowing as measured by sound
transmission. Journal of Prosthetic Dentistry, 46,443-449.
Gomez-Vega, J.M., Hozumi, A., Saiz, E., Tomsia, A.P., Sugimura, H. and Takai, O. (2001).
Bioactive glass-mesoporous silica coatings on Ti6Al4V throgh enameling and triblock-
copolymer-templated sol-gel processing, Journal of Biomedical Materials Research, 56,
382-389.
Gomez-Vega, J.M, Saiz, E., Tomsia, A.P., Marshall, G.W. and Marshall, S.J. (2000).
Bioactive glass coatings with hydroxyapatite and Bioglass particles on Ti-based implants.
Biomaterials, 21, 105-111.
Gomez-Vega, J.M., Saiz, E., Tomsia, A.P., Oku, T., Suganuma, K., Marshall, G.W.,
Marshall, S.J. (2000). Novel bioactive glass functionally graded coatings on Ti6Al4V.
Advanced Materials, 12, 894-898.
Griss, P., Greenspan, D.C., Heimke, G., Krempien, B., Buchinger, R., Hench, L.L. and
Jentschura, G. (1976). Evaluation of a bioglass-coated Al2O3 total hip prosthesis in
sheep. Journal of Biomedical Materials Research, 10, 511-518.
Greenspan, D.C., Hench, L.L. (1976) Chemical and mechanical behavior of bioglass-coated
alumina. J. Biomed. Mater. Res, 10:503-509.
Guldberg, R.E., Richards, M., Caldwell, N.J., Kuelske, C.L. and Goldstein, S.A. (1997).
Mechanical stimulation of tissue repair in the hydraulic bone chamber. Journal of Bone
and Mineral Research, 12, 1295-1302.
Hashimoto, M., Akagawa, Y., Hashimoto, N., Nikai, H. and Tsuru, H. (1988). Single crystal
sapphire endosseous implant loaded with functional stress: clinical and histological
evaluation of peri-implant tissues. Journal of Oral Rehabilitation, 15, 65-76.
Hedia, H.S. (2002). Stress and strain distribution behavior in the bone due to the effect of
cancellous bone, dental implant material and the bone height. Bio-medical Materials and
Engineering, 12, 111-119.
Hench, L.L. (1994). Bioactive ceramics: theory and clinical applications: Ö.H. Andersson, A.
Yli-Urpo (Eds.), Bioceramics 7, (7, pp.3-14). Oxford, UK: Butterworth-Heinmann
publisher.
Hench, L.L., Andersson, Ö.H. (1993). Bioactive glass coatings. In: L.L. Hench, J. Wilson
(Eds.), An introduction to bioceramics (pp. 239), USA: World Scientific.
Hench, L.L., Splinter, R.J., Allen, W.C. and Greenlee, T.K. (1971). Bonding mechanisms at
the interface of ceramic prosthetic materials. Journal of Biomedical Materials Reseach
Symposium, 2, 117-141.
Hodosh, M., Povar, M. and Shklar, G. (1969). The dental polymer implant concept. Journal
of Prosthetic Dentistry, 22, 371-380.
Hodosh, M., Povar, M., Shklar, G. (1970). Control of porosity in polymer tooth implants.
Journal of Prosthetic Dentistry, 24, 453-460.
Hodosh, M., Povar, M., Shklar, G. (1973). Development of the self-supporting polymer tooth
implant. Journal of Biomedical Materials Research, 7, 205-216.
Howell, A.H. and Brudevold, F. (1950). Vertical forces used during chewing of food. Journal
of Dental Research, 29,133-136.
Huiskes, R. (1993). Failed innovation in total hip replacement. Diagnosis and proposals for a
cure. Acta Orthopaedica. Scandinavica, 64, 699-716.
Ichikawa, Y., Akagawa, Y., Nikai, H. and Tsuru, H. (1992). Tissue compatibility and stability
of a new zirconia ceramic in vivo. Journal of Prosthetic Dentistry, 68, 322-326.
Jansen, J.A., von Recum, A.F., van der Waerden, J.P.C.M. and de Groot, K. (1992). Soft
tissue response to different types of sintered metal fiber-web materials. Biomaterials, 13,
959-968.
Jarcho, M. (1992). Retrospective analysis of hydroxyapatite development for oral implant
applications. Dent. Clin. North Am., 36:19-26.
Jones, L.C., Frondoza, C. and Hungerford D.S. (2001). Effect of PMMA particles and
movement on an implant interface in a canine model. Journal of Bone and Joint Surgery
– British Volume, 83, 448-458.
Kangasniemi, I., Verheyen, C., Velde, E and deGroot, K. (1994). In vivo tensile testing of
fluoroapatite and hydroxyapatite plasma-sprayed coatings, Journal of Biomedical
Materials Research, 42, 295-305.
Kaplan, F.S., Hayes, W.C., Keaveny, T.M., Boskey, A., Einhorn J.P. and Iannotti, J.P.
(1994). Form and function of bone. In: S.R. Simon (Ed.), Orthopaedic basic science
(pp.127-184). Rosemont, Illinois: American Academy of Orthopaedic Surgeons.
Katz, J.L. (1995) Mechanics of hard tissue. In: Brozino, J.D., ed. The Biomedical Engineering
Handbook, 273–289. Boca Raton: CRC Press.
Kitsugi, T., Nakamura, T., Oka, M., Senaha, Y., Goto, T. and Shibuya, T. (1996). Bone-
bonding behavior of plasma-sprayed coatings of Bioglass R, AW-glass ceramic, and
tricalcium phosphate on titanium alloy. Journal of Biomedical Materials Research, 30,
261-269.
Klein-Nulend, J., Van der Plas, A., Semeins, C.M., Ajubi, N.E., Frangos, J.A., Nijweide, P.J.
and Burger, E.H. (1995). Sensitivity of osteocytes to biomechanical stress in vitro. The
Journal of the Federation of American Societies for Experimental Biology, 9, 441-445.
Kohal, R.J., Klaus, G. and Strub, J.R. (2006). Zirconia-implant-supported all-ceramic crowns
withstand long-term load: a pilot investigation. Clinical Oral Implants Research, 17, 565-
571.
Krauser, J.T., Boner, C. and Boner, N. (1990). Hydroxyapatite coated dental implants.
Biological criteria and prosthetic possibilities. Les Cahiers de prothèse, 71, 56-75.
Kurtz, S.M. and Devine, J.N. (2007). PEEK Biomaterials in Trauma, Orthopedic, and Spinal
Implants. Biomaterials, 28, 4845-4869.
Lacefield, W.R. (1988). Hydroxlapatite coatings. In: P. Ducheyne, J.E. Lemons, (Eds.),
Bioceramics. materials characteristics vs in vivo behavior, Vol. 523. New York: Annals
of New York Academy of Sciences, 523.
Lacefield, W.R. (1994). Characterization of hydroxylapatite coatings, Journal of Oral

Implantology, 20, 214-220.
Lacefield, W.R. and Hench, L.L. (1986). The bonding of Bioglass to a cobalt-chromium
surgical implant alloy. Biomaterials, 7, 104-108.
Laurell, L. (1985). Occlusal forces and chewing ability in dentitions with cross-arch bridges.
Swedish Dental Journal, 26, 160-7.
Leeuwenburgh, S.C.G, Wolke, J.G.C., Jansen, J.A. and De Groot, K. (2008). Calcium
phosphate coatings. In: T. Kokubo (Ed.), From Bioceramics and Their Clinical
Applications (pp. 464-484). Cambridge, UK: Woodhead Publishing.
Lemons, J.E. (1998) Unanticipated outcomes: dental implants. Implant Dentistry, 7, 351-354.
Li, Z., Kitsugi, T., Yamamuro, T., Chang, Y.S., Senaha, Y., Takagi, H., Nakamura, T. and
Oka, M. (1995). Bone-bonding behavior under load-bearing conditions of an alumina
ceramic implant incorporating beads coated with glassceramic containing apatite and
wollastonite. Journal of Biomedical Materials Research, 29, 1081-1088.
Lum, L.B., Beirne, O.R. and Curtis, D.A. (1991). Histological evaluation of Ha coated vs
uncoated titanium blade implants in delayed and immediately loaded applications.
International Journal of Oral and Maxillofacial Implants, 6, 456-62.
Manley, M., Kotzar, G., Stern, L. and Wilde A. (1987). Effects of repetitive loading on the
integrity of porous coatings. Clinical Orthopaedics and Related Research, 217, 293-302.
Mihalko, W.M., May, T.C., Kajy, J.F. and Krause, W.R. (1992). Finite element analysis of
interface geometry effects on crestal bone surrounding a dental implant. Implant
Dentistry, 1, 212-217.
Mishra, A.K., Davidson, J.A., Poggie, R.A., Kovaac, P. and Fitzgerald, T.J. (1996).
Mechanical and tribological properties of diffusion hardened Ti-13Nb-13Zr—A new
titanium alloy for surgical implants. In: S. Brown, J. Lemons (Eds.), Medical applications
of titanium and its alloys: the material and biological issue (pp. 96-112). West
Conshohocken, PA: American Society for Testing and Materials.
Miyata, T., Kobayashi, Y., Araki, H., Ohto, T. and Shin, K. (2000). The influence of
controlled occlusal overload on peri-implant tissue. Part 3: A histologic study in
monkeys. International Journal of Oral and Maxillofacial Implants, 15, 425-431.
Miyata, T., Kobayashi, Y., Araki, H., Ohto, T. and Shin, K. (2002). The influence of
controlled occlusal overload on peri-implant tissue. Part 4: a histologic study in monkeys.
International Journal of Oral and Maxillofacial Implants. 17, 384-390.
Moritz, N., Rossi, S., Vedel, E., Tirri, T., Ylänen, H., Aro, H. and Närhi, T. (2004). Implants
coated with bioactive glass by CO2-laser, an in vivo study. Journal of Materials Science
Materials in Medicine, 15,795-802.
Mouzin, O., Soballe, K. and Bechtold, J.E. (2001). Loading improves anchorage of
hydroxyapatite implants more than titanium implants. Journal of Biomedical Materials
Research, 58, 61-68.
Mosekilde, L. (2000). Age-related changes in bone mass, structure, and strength-effects of
loading, Zeitschrift fur Rheumatologie, 59, 1–9.
Naert, I., Quirynen, M., van Steenberghe, D. and Darius, P. (1992). A study of 589
consecutive implants supporting complete fixed prostheses. Part II: Prosthetic aspects.
Journal of Prosthetics Dentistry, 68, 949-956.
Nakamura, T., Yamamuro, T., Higashi, S, Kokubo, T. and Itoo, S. (1985). A new glass-
ceramic for bone replacement evaluation of its bonding to bone tissue. Journal of
Biomedical Materials Research, 19, 685-698.
Neill, D.J., Kydd, W.L., Nairn, R.I. and Wilson, J. (1989). Functional loading of the dentition
during mastication. Journal of Prosthetic Dentistry, 62, 218-228.
O‘Brien, W. J. (2002) Dental materials and their selection (3rd edition). Illinois, USA:
Quintessence Publishing.
Ogiso, M., Nakabayashi, N., Matsumoto, T., Yamamura, M. and Lee, R.R. (1996). Adhesive
improvement of the mechanical properties of a dense HA-cemented Ti dental implant.
Journal of Biomedical Materials Research, 30, 109-116.
Oku, T., Suganuma, K., Wallenberg, L.R., Tomsia, A.P., Gomez-Vega, J.M. and Saiz, E.
(2001). Structural characterization of the metal/glass interface in bioactive glass coatings
on Ti-6Al-4V, J. Mat. Sci. Mat. Med., 12:413-417.
Pajamäki, K.J., Lindholm, T.S., Andersson, Ö.H., Karlsson, K.H., Vedel, E., Yli-Urpo, A.
and Happonen, R.P. (1995). Bioactive glass and glass-ceramic-coated hip endoprosthesis:
experimental study in rabbit, Journal of Materials Science Materials in Medicine, 6; 14-
18.
Paldan, H., Areva, S., Tirri, T., Peltola, T., Lindholm, T.C., Lassila, L., Pelliniemi, L.J.,
Happonen, R.P. and Närhi, T.O. (2008). Soft tissue attachment on sol–gel-treated
titanium implants in vivo. Journal of Materials Science Materials in Medicine, 19, 1283-
1290.
Paphangkorakit, J. and Osbom, J.W. (1987). The effect of pressure on maximum incisal bite
force in man. Archives of Oral Biology, 2, 11–7.
Piattelli, A., Ruggierie, A., Franchi, M., Romasco, N. and Trisi, P. (1993). A histologic and
histomorphometric study of bone reactions to unloaded and loaded non-submerged single
implants in monkeys: a pilot study. Journal of Oral Implantology, 19, 314-320.
Piattelli, A., Paoloantonio, M., Corigliano, M. and Scarano, A. (1997). Immediate loading of
titanium plasma-sprayed screw-shaped implants in man: a clinical and histological report
of 2 cases. Journal of Periodontology, 68, 591-597.
Piconi, C., Burger, W., Richter, H.G., Cittadini, A., Maccauro, G., Covacci, V., Bruzzese, N.,
Ricci, G.A. and Marmo, E. (1998). Y-TZP ceramics for artificial joint replacements.
Biomaterials, 19, 1489-1494.
Piotrowski, G., Hench, L.L., Allen, W.C. and Miller, G.J. (1975). Mechanical studies of the
bone Bioglass interfacial bond, Journal of Biomedical Materials Research, 6, 47-61.
Prendergast, P.J. (1997). Finite element models in tissue mechanics and orthopaedic implant
design. Clinical Biomechanics, 12, 343-366.
Quirynen, M., Naert, I., van Steenberghe, D. and Nys, L. (1992). A study of 589 consecutive
implants supporting complete fixed prostheses. Part I: Periodontal aspects. Journal of
Prosthetic Dentistry. 68, 655-663.
Rangert, B., Jemt, T. and Jorneus, L. (1989). Forces and moments on Brånemark implants.
International Journal of Oral and Maxillofacial Implants, 4, 241-247.
Rangert, B., Krogh, P.H., Langer, B. and Van Roekel, N. (1995). Bending overload and
implant fracture. International Journal of Oral and maxillofacial Implants, 10, 326-334.
Richter, H.G., Burger, W. and Osthues, F. (1994). Zirconia for medical implants-the role of
strength properties. In: Ö.H. Andersson, A. Yli- Urpo (Eds.), Bioceramics 7 (pp. 401-
406). London: Butterworth-Heinemann Publishers.
Richter, E.J. (1998). In vivo horizontal bending moments on implants. International Journal
of Oral and Maxillofacial Implants, 13, 232-244.
Rickerby, D.S. (1988). Review of the methods for measurement of coating-substrate
adhesion. Surface and Coating Technology, 36, 541-557.
Rossi, S., Moritz, N., Tirri, T., Peltola, T., Areva, S., Jokinen, M., Happonen, R.P. and Närhi,
T. (2007). Comparison between sol-gel-derived anatase- and rutile-structured TiO2
coatings in soft-tissue environment. Journal of Biomedical Materials Research Part A,
82, 965-974.
Rubin, M.A., Rubin, J. and Jasiuk, I. (2004). SEM and TEM study of the hierarchical
structure of C57BL/6J and C3H/HeJ mice trabecular bone. Bone, 35, 1-20.
Sagara, M., Akagawa, Y., Nikai, H. and Tsuru, H. (1993). The effects of early occlusal
loading on one-stage titanium implants in beagle dogs: a pilot study. Journal of
Prosthetic Dentistry, 69, 281-288.
Saiz, E., Goldman, M., Gomez-Vega, J.M., Tomsia, A.P., Marshall, G.W. and Marshall, S.J.
(2002). In vitro behaviour of silicate glass coatings on Ti6Al4V. Biomaterials, 23, 3749-
3756.
Schepers, E., Ducheyne, P. and De Clerq, M. (1989). Interfacial analysis of fiber-reinforced
bioactive glass dental root implants, Journal of Biomedical Materials Research, 23, 735-
752.
Schrooten, J. and Helsen, J.A. (2000). Adhesion of bioactive glass coating to Ti6A14V oral
implant. Biomaterials, 21, 1461-1469.
Schrooten, J., Van Oosterwyck, H., Vander Sloten, J., Helsen, J.A. (1999). Adhesion of new
bioactive glass coating. Journal of Biomedical Materials Research, 44, 243-252.
Schulte, W. (1984). The intra-osseous Al2O3 (Frialit) Tuebingen Implant. Developmental
status after eight years (I-III). Quintessence International, 15, 1-39.
Schulte, W. (1995). Implants and the periodontium. International Dental Journal, 45:16-26.
Skerry, T.M., Bitensky, L., Chayen, J. and Lanyon, L.E. (1989). Early strain-related changes
in enzyme activity in osteocytes following bone loading in vivo. Journal of Bone and
Mineral Research, 4, 783-788.
Skripitz, R. and Aspenberg, P. (2001). Early effect of parathyroid hormone (1–34) on implant
fixation. Clinical Orthopaedics and Related Research, 392, 427-432.
Soballe, K. (1993) Hydroxyapatite ceramic coating for bone implant fixation, Acta
Orthopaedica Scandinavica, 64, 1-58.
Smedberg, J.I., Nilner, K., Rangert, B., Svensson, S.A. and Glantz, S.A. (1996). On the
influence of superstructure connection on implant preload: a methodological and clinical
study. Clinical Oral Implants Research, 7, 55-63.
Suominen, H. (1993). Bone mineral density and long term exercise. An overview of cross-
sectional athlete studies. Sports Medicine, 16, 316-330.
Szmukler-Moncler, S., Salama, S., Reingewirtz, Y., Dubruille, J.H. (1998). Timing of loading
and effect of micro-motion on bone–implant interface: a review of experimental
literature. Journal of Biomedical Materials Research, 43, 193-203.
Toksvig-Larsen, S., Jorn, L.P., Ryd, L. and Lindstrand A. (2000). Hydroxyapatite-enhanced
tibial prosthetic fixation. Clinical Orthopaedics and Related Research, 370, 192–200.
Van Eijden, T.M. (1991). Three-dimensional analyses of human bite-force magnitude and
moment. Archives of Oral Biology, 36, 535-539.
Van Noort R. (2002). Introduction to dental materials (2nd edition). London, UK: Mosby.
Van Staden, R.C., Guan, H. and Loo, Y.C. (2006). Application of the finite element method
in dental implant research. Computer Methods in Biomechanics and Biomedical
Engineering, 9, 257-270.
Vico, L., Collet, P., Guignandon, A., Lafage-Proust, M.H., Thomas, T., Rehaillia, M. and
Alexandre, C. (2000). Effects of long-term microgravity exposure on cancellous and
cortical weight-bearing bones of cosmonauts. Lancet 355, 1607-1611.
Waltimo, A. and Kononen, M. (1993). A novel bite force recorder and maximal isometric bite
force values for healthy-young adults. Scandinavian Journal of Dental Research, 101,
171-175.
West, J.K., Clark, A.E., Hall, M.B. and Turner, G.E. (1990). In vivo bone-bonding study of
Bioglass-coated titanium alloy. In: T. Yamamuro, L.L. Hench, J. Wilson (Eds.),
Handbook of bioactive ceramics (pp. 161-166). Boca Raton FL: CRC Press.
White, S.N., Caputo, A.A. and Anderkvist, T.A. (1994). Effect of cantilever length on stress
transfer by implant-supported prostheses. Journal of Prosthetic Dentistry, 71, 493-499.
Wilson, J., Schoen, F.J., Pigott, G.H. and Hench, L.L. (1981). Toxicology and bioactivity of
Bioglasses. Journal of Biomedical Materials Research, 15, 805-817.
Wilson, J. and Noletti, D. (1990). Bonding of soft tissues to Bioglass. In: T. Yamamuro, L.L.
Hench, J. Wilson (Eds.), Handbook of bioactive ceramics (pp. 235-244). Boca Raton:
CRC Press.
Wintermantel, E. and Mayer, J. (1995). Anisotropic biomaterials strategies and developments
for bone implants. In: D.L. Wise, D.J. Trantolo, D.E. Altobelli, J.D. Yaszemiski, J.D.
Gresser, and E.R. Schwartz (Eds.), Encyclopedic Handbook of Biomaterials and
Bioengineering (1st, pp. 3-42). NewYork: Mearcel Dekker.
Wolff, J. (1892). Das Gesetz der Transformation der Knochen. Berlin A. Hirchwild.
Translated as: The Law of Bone Remodeling. Maquet P. and Furlong R. (1986). Berlin:
Springer-Verlag.
Wolke, J.G.C., deBlieck-Hogervorst, J.M.A., Dhert, W.J.A., Klein, C.P.A.T. and deGroot, K.
(1992) Studies on the thermal spraying of apatite bioceramics, Journal of Thermal Spray
Technology, 1, 75-85.
Yaszemski, M.J., Payne, R.G., Hayes, W.C., Langer, R., and Mikos, A.G. (1996). Evolution
of bone transplantation: molecular, cellular, and tissue strategies to engineer human bone.
Biomaterials. 17, 175-185.
Zerwekh, J.E., Ruml, L.A., Gottschalk, F. and Pak, C.Y.C. (1998). The effects of twelve
weeks of bed rest on bone histology, biochemical markers of bone turnover, and calcium
homeostasis in eleven normal subjects. Journal of Bone and Mineral Research, 13, 1594-
1601.
Zhao, D.S., Moritz, N., Laurila, P., Mattila, R., Lassila, L.V.J., Strandberg, N., Mäntylä, T.,
Vallittu, P.K. and Aro, H.T. (2009). Development of a Multi-Component Fiber-
Reinforced Composite Implant for Load-Sharing Conditions. Medical Engineering and
Physics, 31, 461-469.
Chapter 4
BIOMECHANICAL REMODELING OF THE DIABETIC

GASTROINTESTINAL TRACT
Jingbo Zhao, Donghua Liao, Jian Yang and Hans Gregersen

Mech-sense, Aalborg Hospital Science and Innovation Centre (AHSIC),
9000 Aalborg, Denmark
ABSTRACT
Gastrointestinal tract sensory-motor abnormalities are common in patients with
diabetes mellitus with symptoms arising from the whole GI tract. Common complaints
include dysphasia, early satiety, reflux, constipation, abdominal pain, nausea, vomiting,
and diarrhea. The pathogenesis of GI symptoms in diabetes mellitus is complex in nature,
multi-factorial (motor dysfunction, autonomic neuropathy, glycemic control,
psychological factors, etc.) and is not well understood. Histologically, many studies have
demonstrated prominent proliferation of different GI wall layers during diabetes. During
the past several years, several studies demonstrated that experimental diabetes induces GI
morphological and biomechanical remodeling. Following the development of diabetes,
the GI wall becomes thicker and the stiffness of the GI wall increases in a time-dependent
manner. It is well known that mechanosensitive nerve endings exist in the GI tract where
they serve a critical role for tissue homeostasis and symptom generation.
Mechanoreceptor-like structures such as intraganglionic laminar nerve endings and
intramuscular arrays have been identified. The changes of stress and strain in the GI wall
will alter the biomechanical environment of the mechanosensitive nerve endings,
therefore, the structure as well as the tension, stress and strain distribution in the GI wall
is important for the sensory and motor function. Biomechanical remodeling of diabetic
GI tract including alterations of residual strain and increase in wall stiffness will alter the
tension and stress distribution in the vicinity of the mechanosensitive afferents with
consequences for perception and motility of the GI tract.
Keywords: Diabetes, gastrointestinal tract, mechanosensory, biomechanics, remodeling.

Telephone: +45 99326907, E-mail: jz@rn.dk, jzhao@hst.aau.dk
138 Jingbo Zhao, Donghua Liao, Jian Yang et al.
INTRODUCTION
Diabetes mellitus (DM) is a chronic disease requiring lifelong medical attention in order
to limit the development of potentially devastating late complications and to manage them if
they occur. DM affects almost 6% of the world's population, means that approximately 246
million people worldwide were affected (International Diabetes Federation
[http://www.idf.org/]) and the number of diabetic patients will reach 300 million in 2025
[Adeghate et al., 2006]. Therefore, it is becoming one of the major public health problems
because a great proportion of the healthcare expenditure has been spent on the treatment of its
associated morbidity and mortality. In addition, persons with DM have a lower quality of life
compared with persons without DM.
Chronic complications of DM include nerve disease, heart disease, vascular disease,
kidney disease, eye disease as well as digestive disease [Nair, 2007]. Gastrointestinal (GI)
disorders are common in diabetic patients [Verne and Sninsky, 1998; Folwaczny et al., 1999].
The entire GI tract from the esophagus to the anorectal region can be affected. The
pathogenesis of the GI abnormalities is complex of nature, multi-factorial (motor dysfunction,
autonomic neuropathy, glycemic control, psychological factors, etc.) and is not well
understood [Horowitz and Samsom, 2004]. A number of motor dysfunctions have been
described in different segments of the GI tract in diabetic patients: esophagus (dysmotility),
stomach (dysmotility, delayed emptying) and small and large bowel (dysmotility, delayed
transit, bacterial overgrowth, and diarrhea) [Horowitz and Samsom, 2004]. A few studies
have addressed the visceral sensory function in DM [Kamath et al., 1998, 2000; Rathmann et
al., 1991]. Although the hyperglycemia and neuropathy seem to be the main mechanisms to
the motor-sensory dysfunction in the GI tract in DM, many studies have however
demonstrated morphological changes [Mayhew et al., 1989; Tahara and Yamamoto, 1988;
Zoubi et al., 1995; Charlton et al., 2000] and biomechanical remodeling in the GI tract in DM
[Jorgensen et al., 2001; Liao et al., 2006; Sha et al., 2006; Yang et al., 2004, 2006; Zeng et al.,
2004; Zhao et al., 2002, 2003a, 2003b, 2006, 2007, 2008; Frokjaer et al., 2007a, 2007b].
Therefore, the morphological and biomechanical remodeling of the GI wall may also be an
important factor in the pathogenesis to the GI motor-sensory dysfunction in the diabetic
patients. Data on the biomechanical properties are crucial for the understanding of the motor
function of the GI tract because, 1) peristaltic motion that propels the food through the GI
tract is a result of interaction of the passive and active tissue forces and the hydrodynamic
forces in the food bolus and 2) remodeling of the mechanical properties reflects the changes
in the tissue structure that determine a specific motor dysfunction.
SYMPTOMS FROM GI TRACT IN DM

Motor and sensory abnormalities in DM can affect the entire GI tract or part hereof, and
the perceived symptoms may originate from one or several parts of the GI tract. The
prevalence of GI symptoms is high in both insulin dependent DM (IDDM) and non-insulin
dependent DM (NIDDM) [Ko et al., 1999; Maleki et al., 2001; Ricci et al., 2000; Schvarcz et
al., 1996; Spangeus et al., 1999; Talley et al., 2001, 2002]. The symptoms relating to the
esophagus, stomach, small intestine and colon are summarized in Table 1. Most symptoms
Biomechanical Remodeling of the Diabetic Gastrointestinal Tract 139
are non-specific of nature, therefore when dealing with GI symptoms; some specific issues
need to be addressed. Chest pain may relate to reflux or esophageal motor disorders, but
ischemic heart disease and other causes of non-cardiac chest pain are also possible causes and
must be excluded. Dysphagia, the most characteristic symptom of impaired esophageal
transit, may be caused by motility disorders of the esophagus. A number of other incidental
conditions must be excluded when diabetic gastroparesis is suspected: gastric outlet
obstruction caused by tumors and ulcer disease; metabolic abnormalities such as diabetic
ketoacidosis or uremia and side effects of pharmacotherapy. Diabetic patients frequently
report abdominal pain and this may be the only symptom of diabetic gastroparesis; however,
abdominal pain can also be seen in diabetic ketoacidosis and severe metabolic acidosis
[Umpierrez and Freire, 2002]. Diabetic patients with thoracic polyradiculopathy, a rare
condition, may also suffer from abdominal pain [Longstreth, 1997]. Diarrhea in DM patients
may be induced by a number of factors. These may include food composition, abnormal
intestinal motility, small intestinal bacterial overgrowth, excessive loss of bile acids,
pancreatic insufficiency and more [Samsom and Verhagen, 2004]. Abnormal small intestinal
motility (rapid or delayed transit) is a frequent condition in the diabetic patients as described
below. Rapid transit may induce an increase in intra-luminal contents that reach the caecum,
whereas delayed transit may cause bacterial overgrowth, both potentially resulting in diabetic
diarrhea. Celiac disease is overrepresented in IDDM and a cause of severe diarrhea to be
excluded when dealing with diabetic diarrhea [Anderson, 2005]. Disordered defecation,
characterized by fecal incontinence and constipation occurs frequently in patients with DM
[Sun and Read, 2004]; the patients frequently fail to report disturbances in defecation unless
the symptoms are severe. Constipation has been cited as the most common gastrointestinal
complaint of diabetic patients [Batte et al., 1983]. Profound constipation with massive fecal
impaction may occur [Batte et al., 1983]. These patients may develop marked abdominal
distention and severe nausea and vomiting associated with electrolyte disturbances. Prolonged
fecal impaction may lead to stercoral ulcerations and perforation. Steatorrhea has generally
been attributed to pancreatic insufficiency, celiac disease, or bacterial overgrowth of the small
intestine [Batte et al., 1983]. In these patients the colon appears to play a secondary or
permissive role in the production of diarrhea. In the absence of steatorrhea, colonic
dysfunction may be a primary contributor to diabetic diarrhea.
Table 1. Symptoms of GI tract in DM
Organs Symptoms
Esophagus Heartburn, dysphagia and chest pain
Stomach Postprandial nausea, vomiting, early satiety, bloating, weight loss
and abdominal pain
Small intestine Diarrhea, discomfort, pain and pseudo-obstruction
Large intestine Faecal Incontinence, constipation, diarrhea
MECHANO-SENSORY DYSFUNCTION OF GI TRACT IN DM

Many studies have demonstrated that DM patients have disorders of transit and motility.
However, the data related to the sensory dysfunction of the GI tract are sparse compared with
those relating to the motor dysfunction of the GI tract. The most frequent motility disorders of
GI tract are shown in Table 2.
Esophagus: Impaired esophageal transit has been reported both in IDDM and NIDDM
patients [Horowitz et al., 1991; Jorgensen et al., 1991; Tsai et al, 1995]. The esophageal
transit appears to be delayed in about 50% of patients with long-standing DM [Annese et al.,
1999]. The retarded esophageal transit in the DM usually reflects either peristaltic failure or
focal low-amplitude pressure waves [Holloway et al., 1999]. Esophageal manometric
abnormalities occur in over 50% of patients with DM [Smout, 2004] (see Table 2). The
reduced amplitude of lower esophageal sphincter pressure is in accordance with the increased
prevalence of gastroesophageal reflux disease in DM. More recently the evoked esophageal
contractile activity to standardized bag distension was assessed using a specialized
ultrasound-based probe by Frokjaer et al [2007]. A balloon-like bag was positioned 10 cm
above the lower esophageal sphincter and inflated. It was demonstrated, both at the bag and 6
cm proximally, that the distension induced hyperreactivity and impaired coordination of the
contractions in the diabetic patients.
Elevation of perception thresholds to esophageal electrical stimulation has been observed
in patients with DAN and different severity of GI symptoms [Rathmann et al., 1991].
Increased vagal tonus and abnormal evoked brain potentials to mechanical and electrical
stimulation of the esophagus has also been shown [Kamath et al., 1998].
Table 2. Motility disorders of GI tract in DM
Organs Motility disorders

Esophagus 1) amplitude and number of peristaltic contractions↑
2) number of spontaneous and non-propagated contractions↑
3) amplitude of lower esophageal sphincter pressure↓
4) multi-peaked contractions
Stomach 1) antral IMMC↓
2) post-prandial antral activity and the number of antral
contractions↓
3) pyloric dysmotility
Small intestine 1) frequency and amplitude of the antropyloroduodenal
contractions↑↓
2) duration of MMC cycle↑
3) early recurrence of the MMC and clusters of contractile activity
Colon 1) enhanced spontaneous activity
2) Increase in stretch-induced rhythmic motor activity
Stomach: Several animal studies reported a slowing gastric emptying in IDDM and
NIDDM rats [Mehta et al., 2002; Ogata et al., 1997; Liu et al., 2004], whereas other studies
on IDDM and NIDDM animals demonstrated that gastric emptying increased [Nowak et al.,
1994; Ogata et al., 1996; Young et al., 1995]. Using radionuclide measuring techniques it has
been demonstrated that gastric empting of solid, or a liquid meal was abnormally slow in 30-
50% of patients with long-standing IDDM and NIDDM [Horowitz et al., 1996]. Gastroparesis
in DM has been known clinically for more than 50 years [Smith and Ferris, 2003]. The gastric
emptying delay in DM is related to both slow transit with increased retention of food in the
proximal and distal stomach [Jones et al., 1995], and abnormal motility of the gastric wall
[Samsom et al., 1998]. Motility disorders of the fundus and pylorus have been demonstrated
in the diabetic animals [James et al., 2004; Korolkiewicz et al., 1998]. In human studies, it is
recognized that disordered gastric contractile activity as assessed by manometry and gastric
emptying occurs frequently in DM [Bjornsson et al., 1994; Samsom et al., 1996]. The motility
disorders may include three aspects: Inter-digestive migrating motor complex (IMMC),
amplitude and frequency of contractions, and pyloric dysfunction (Table 2).
Rayner et al [2000] did isovolumetric and isobaric distensions of the proximal stomach in
ten randomly selected patients with IDDM. They demonstrated that the perception of gastric
distension during euglycemia was increased compared with healthy controls. To study
mechanisms behind postprandial symptoms in patients with diabetes, the gastric
accommodation of the meal was assessed by abdominal ultrasound [Undeland., 1998]. In DM
patients, a large proximal stomach was associated with perception of fullness and a large
antrum was associated with perception of pain after a meal.
Small intestine: Delayed and rapid transit in the small intestine was observed in animal
diabetic models. The GI transit was rapid in non-obese diabetic mice [el Salhy, 2001] and was
slower in obese diabetic mice [el Salhy, 2002]. Anjaneyulu and Ramarao [2002] reported an
increase in intestinal transit and a decrease in intestinal tone due to increased cholinergic and
decreased beta-adrenergic receptor activities in DM rats. Slow small intestinal transit in DM
patients have been documented using breath hydrogen appearance time after the ingestion of
lactulose [de Boer et al., 1993], by using radiopaque markers [Iida et al., 2000] and by use of
metal-detector test [Folwaczny et al., 1995]. On the other hand, Keshavarzian and Iber [1986]
investigated intestinal transit in IDDM patients after the ingestion of both liquid and solid
meals and showed abnormal fast intestinal transit in their sample of diabetic patients. Nguyen
et al [1997] used intraluminal multiple impedance measurements to identify the postprandial
duodenal chyme transport in patients with long standing IDDM. They demonstrated that the
patients had disturbed propulsive chyme transport through the duodenum and the duodenal
chyme clearance activity was decreased. Camilleri and Malagelada [1984] reported that small
intestinal motility was abnormal in about 80% patients of long-standing DM with delayed
gastric emptying. Both postprandial and fasting small intestinal dysmotility in the DM was
reported (Table 2). Dooley et al [1988] studied fasting GI motility by manometry for a mean
of 210 min in a group of 12 NIDDM patients with diarrhea and DAN. The patients showed
grossly disordered motility. The migrating motor complex (MMC) disorders reflect the
prolongation of phase II without change in the duration of phase I and III. The results of
studies on postprandial motility at the level of the small intestine are inconsistent. However,
abnormal motility patterns were observed in the diabetic subjects [Samsom et al., 1996].
More recently, Frøkjær et al [2007] used a multimodal stimulation device to investigate
the visceral sensitivity to mechanical, thermal and electrical stimulation in the esophagus and
duodenum in IDDM patients with DAN and GI symptoms. This study demonstrated that the
patients had decreased sensitivity to the stimulations of the esophagus and duodenum. This
indicates that the affection of the sensory nerves is widespread in the GI tract. As the
multimodal approach is thought to stimulate the mucosa, submucosa and muscle layers
differentially, the disease seems to be generalized to nerves in all layers of the GI tract.
Colon: Delayed transit was frequent in males with IDDM of greater than 10 years
duration and the symptoms related closely to the region impaired [Iber et al., 1993]. The mean
total colonic transit time of the diabetic patients was significantly longer than that in healthy
subjects. Furthermore, the diabetic patients with constipation showed longer total, left and
recto-sigmoid colonic transit times than those without constipation [Jung et al., 2003].
Diabetic patients with mild constipation had a postprandial increase in colonic motility;
however, the response was delayed to 60-90 min after eating. Diabetic patients with severe
constipation had no postprandial increase in colonic motility. These studies suggest that
patients with diabetes mellitus and severe constipation may have a DAN which leads to an
absent postprandial gastrocolonic response [Battle et al., 1980]. A marked alteration of
motility in the colon was also observed in diabetic animals; however, the change in motility is
unlikely to be due to a change in inhibitory control mechanisms and may be due to an
increased excitability [Forrest and Parsons, 2003]. In diabetic rats, the stretch-induced
rhythmic contractile activity remained robust and of similar frequency but was significantly
higher in amplitude compared with that in control rats. This robust presence of low-frequency
contractions is consistent with the unaffected pacemaker, the ICC associated with Auerbach's
plexus, and the increased amplitude correlates with loss of and injury to ICC of the
submuscular plexus and intramuscular ICC [Forrest et al., 2008]. Hyperglycemia blunts
mechanoreceptor-mediated gastrocolonic responses and ascending contractions but not
descending components of the peristaltic reflex in humans, effects not caused by
hyperinsulinemia or direct muscle actions. These inhibitory effects on long and short neural
reflexes that modulate colonic motility may contribute to constipation in diabetes [Sims et al.,
1995].
MORPHOLOGICAL AND BIOMECHANICAL REMODELING OF GI

TRACT IN DIABETES MELLITUS
Although the hyperglycemia and neuropathy seem to be the main mechanisms to the
motor-sensory dysfunction in the GI tract in DM, the question remains whether the disordered
motor and sensory functions of the GI tract are only due to the neuronal changes and
dysfunction or if primary diabetes-induced structural and biomechanical changes in the GI
tract also play a role? Histologically, many studies have demonstrated prominent proliferation
of different GI wall layers during diabetes.
During the past several years, several studies demonstrated that experimental diabetes
induces GI morphological and biomechanical remodeling. Following the development of
diabetes, the GI wall becomes thicker and the stiffness of the GI wall increases in a time-
dependent manner. Therefore, the morphological and biomechanical remodeling of the GI
wall may also be an important factor in the pathogenesis to the GI motor-sensory dysfunction
in the diabetic patients.
Morphological and Biomechanical Remodeling of Different Parts in GI
Esophageal remodeling: Many studies have shown that DM causes morphological

changes and biomechanical remodeling in the esophagus. Yang et al [2004] in the in vitro
study on the STZ-induced diabetic rat esophagus found that the wall thickness and cross-
sectional wall area increased after the induction of diabetes. During development of the
diabetes, the esophagus became stiffer in the circumferential (Figure 1A) and the longitudinal
as well as in the shear directions [Yang et al., 2004].
Figure 1. A: Effect of diabetes on the circumferential stress and strain of intact oesophageal wall during
the 4 weeks study period. Diabetes shifted the curve to the left indicating increased circumferential
stiffness. B: Circumferential stress-strain relations in the muscle layer and mucosa-submucosa layer at a
longitudinal stretch ratio of 1.5. The curve for the mucosa-submucosa layer during diabetes shifted to
the left in the circumferential direction compared to the control group indicating that the mucosa-
submucosa layer became stiffer in the diabetic rats. N: Normal control, D: 28 days diabetes.
Figure 2. A: The wall thickness in the no-load state and zero-stress state were significantly bigger in the
GK rats than in the normal rats (P<0.01). B: The collagen area fraction in the mucosa-submucosa layer
of the esophagus in the GK rats was significantly higher than that in the normal rats (P<0.01). C: The
opening angle was significantly smaller in the GK rats than in the normal rats (P<0.01). D: The
circumferential stress-strain curve in the GK rats was shifted to the left, indicating that the esophageal
wall stiffness increased in the GK rats.
Detail layered wall biomechanical analysis showed that the morphometric and
biomechanical remodeling in the mucosa-submucosa layer of the twenty-eight days diabetic
esophagus was characterized by increased circumferential length, luminal area and
circumferential stiffness (Figure 1B) [Yang et al., 2006]. The esophageal remodeling was also
found in NIDDM rat esophagus [Zhao et al., 2007]. In the diabetic rats the esophageal weight
per length, wall thickness (figure 2A), cross-sectional wall area and the collagen fraction
(figure 2B) in the mucosa-submucosa layer increased in the NIDDM rats. The opening angle
(figure 2C) and residual strain distribution in the outer surface of the wall decreased, whereas
the passive circumferential stiffness of the esophageal wall increased (figure 2D). One
previous human study reported that among six cases of diffuse muscular hypertrophy of the
esophagus, four cases were associated with DM [Iyer et al., 1986]. Recently, Frøkjær et al
[2007] in a human study found an increase in esophageal wall thickness and altered
deformation to the distension with reduced longitudinal shortening and the radial stretch in
IDDM patients.
Stomach remodeling: A morphological study in DM rats demonstrated that the gastric
mucosa thickness increased in DM rats compared with controls [Watanabe et al., 1995].
Remodeling of the interstitial cells network of Cajal in the stomach were found both in
animals and humans with DM [Forster et al., 2005; Horvath et al., 2005; Long et al., 2004]. A
histopathological study of the human stomach in DM patients with severe gastroparesis
showed prominent collagenization and smooth muscle atrophy of the muscle layer [Ejskjaer
et al., 1999]. Regarding the biomechanical remodeling of the stomach only one report by Liao
et al is available [Liao et al., 2006]. The rat stomach was distended in vitro. Gastric
compliance, the surface tension, and circumferential and longitudinal deformation-pressure
curves were calculated based on three-dimensional ultrasound reconstructions of non-diabetic
and diabetic stomach models. In experimental DM, gastric compliance was lowered both in
the non-glandular stomach (proximal part) and the glandular stomach (distal part) (figure 3).
Furthermore, the circumferential stiffness in the non-glandular part increased (figure 3). The
structural changes of the stomach due to DM may together with the sensory and motor nerve
dysfunction contribute to the delayed gastric emptying and the symptoms in diabetic patients.
Small intestinal remodeling: Many human [Folwaczny et al., 1999; Charlton et al., 2000;
He et al., 2001; Secondulfo et al., 2004] and animal [Folwaczny et al., 1999; Verne and
Sninsky, 1998; Zoubi et al., 1995; Bhor et al., 2004; Noda et al., 2001; Tormo et al., 2002;
Adachi et al., 2003] studies have shown that DM causes morphological alterations in the
small intestine. However, only few data exist on biomechanical remodeling of the small
intestine in DM [Jorgensen et al., 2001].
Recently we did a series of studies on the morphological and biomechanical remodeling
of small intestine in the STZ-induced diabetic rats [Sha et al., 2006; Zhao et al., 2002, 2003a,
2003b]. The major findings were that the opening angle and residual strain were lower in the
duodenum and larger in the jejunum and ileum in diabetic rats compared to normal rats
(Figure 4).
Figure 3. Left: Non-dimensional surface area-pressure curve for the glandular (top) and the non-
glandular stomach part (bottom) in the non-diabetic (solid lines) and the diabetic (dashed lines) rats.
Surface area for the non-glandular stomach in the non-diabetic group increased more than the diabetic
group (P<0.02). Right: The circumferential stretch ratio-pressure for the glandular (top) and the non-
glandular part (bottom) in the non-diabetic (solid lines) and the diabetic (dashed lines) rats. The
circumferential deformation of the non-glandular part in the diabetic group was smaller than that in the
non-diabetic group (P<0.01).
Figure 4. The opening angles and residual strain of the different segments of the intestine in the STZ-
induced diabetic rats. The opening angle did not change significantly in the first two weeks in the
duodenal and jejunal segments. Thereafter it decreased in the duodenum and increased in the jejunum.
The opening angle was increased after 4 days diabetes in the ileum. The same patterns as the opening
angle were found for the residual strain at the mucosa and serosa.
Stress-strain relationship analysis showed that both in the circumferential and

longitudinal directions the stiffness of the intestinal wall increased with the duration of
diabetes (Figure 5). The viscoelastic behavior of intestinal wall also changed during the
development of diabetes (Figure 6). Furthermore, we confirmed previous data that the
intestinal weight, weight per-unit length and wall thickness and area increased in the diabetic
rats (Figure 7).The stress-strain distribution and viscoelastic behavior mainly reflects the
elastic properties of intestine. The changes in elastic properties reflect the structural
remodeling of the intestinal wall during the diabetic development.
Figure 5. The circumferential (top) and longitudinal (bottom) stress-strain relations of the duodenum a
in the STZ-induced diabetic rats. The stress-strain curves in both directions shifted to the left during
experimental diabetes indicating the wall became stiffer.
Figure 6. The mean reduced relaxation function vs linear time scale in the time period of 600 seconds
(top) and 10 seconds (bottom) was shown. The curves appear in the order of largest-to-smallest G (t) as
W4, W1, 4d and Normals. The stress relaxation decreased with the development of experimental
diabetes (all correlation coefficients > 0.99).
Figure 7. The micro-photographs show the normal (left) and 4 weeks diabetic (right) duodenal
histological sections. It clearly demonstrated that the muscle and submucosa layers in the diabetic
duodenum became much thicker than in the normal duodenum. The bar is 100 um.
Colon remodeling: Some studies have demonstrated the morphological and histological
remodeling in the colon due to DM, for example, 1) changes in the innervation of the colon of
the streptozotocin-diabetic model shortly after the induction of diabetes [Lincoln et al., 1984];
2) abnormalities of endocrine cells in the colon of type 2 diabetic animal model [Spangeus et
al., 1998, 1999]; 3) impairment of nitrergic enteric neurons in spontaneously diabetic rats
[Yoneda et al., 2001]; 4) deficiency of KIT-positive cells in the colon of patients with DM
[Nakahara et al., 2002]; 5) early activated the apoptosis cascade in DAN patients [Guo et al.,
2004]; 6) reduced density of ghrelin-immunoreactive cells in animal models of human
diabetes type 1 and 2 [Rauma et al., 2006]; and 7) increased the thickness of the subepithelial
collagen layer in the patients with DM [Unal et al., 2008]. However, it lacks the data of
biomechanical remodeling in the colon due to DM. Recently we investigated the time-
dependent biomechanical and morphometric properties of the colon in diabetic rats [Zhao et
al., 2008]. The colon weight per-unit length, wall thickness and area, and the thickness of
different layers increased in the diabetic rats. For most parameters the biggest changes was
observed during the first four weeks where after it leveled off. The opening angle and residual
strain were bigger in the colon in diabetic rats compared to normal rats. Stress-strain analysis
showed that both in the circumferential and longitudinal directions the stiffness of the colon
wall increased with the duration of diabetes (Figure 8).
Figure 8. The top figure shows the no-load state and zero-stress state in control, 4W and 8W
streptozotocin-induced diabetic rats, the wall thickness and opening angle increased during the
development of diabetes. The bottom figure shows the relation between circumferential (left) and
longitudinal (right) stress and strain. Both in the circumferential direction and the longitudinal
directions, the stress-strain curves shifted to the left in the 4W and 8W diabetic groups compared to
those in the control group. Thus, the colon wall stiffness increased in both directions during the
development of diabetes.
Mechanisms of Morphological and Biomechanical Remodeling of GI Tract in

Diabetes
Untreated diabetic rats eat 2-3 times as much as controls, and it has been suggested that
this hyperphagia account for the observed GI growth seen in these animals [Jervis and Levin
RJ, 1966]. However, diabetic rats‘ pair-feed diets isocaloric with those of controls still exhibit
significantly greater small bowel mass and enhanced crypt cell DNA synthesis [Miller et al.,
1977; Granneman and Stricker, 1984]. This indicates that at least part of the growth response
of GI adaptation in diabetes is independent of increased nutrient consumption. The
mechanism by which hyperphagia stimulates GI growth is unclear. One study demonstrated
increased concentrations of both plasma and tissue glucagon-like peptide 2 (GLP-2) that
correlate with GI growth and adaptation to STZ diabetes. Furthermore, both the elevation in
GLP-2 levels and the bowel growth response were reverted by insulin therapy. Interestingly,
the increments in plasma levels of GLP-2 preceded any changes in intestinal weight, thereby
providing further evidence for a relationship between GLP-2 and the induction of GI growth
in diabetes [Fischer et al., 1997]. Their study supports the idea that an increased nutrient load
may stimulate release of hormones trophic to the GI tract. The combination of hyperphagia
and increased gastric emptying [Granneman and Stricker, 1984] observed in diabetic rats
results in a greater nutrient load being delivered to the small intestine. Because luminal
nutrients, including fat and carbohydrate, are physiological L cell secretagogues in vivo
[Brubaker et al., 1991; Vahl and D'Alessio, 2004] , this provides a possible mechanism
whereby increased luminal nutrients stimulate release of GLP-2 into the circulation, thereby
exerting a trophic action on the intestinal epithelium. It is also well known that the epithelial
homeostasis is balanced by regulation of cell proliferation and cell death. The active form of
caspase-3 was attenuated 1 week after streptozotocin treatment. Suppression of apoptosis in
the early days of STZ-induced diabetes was responsible for the increased mucosal height in
the small intestine in STZ-induced diabetic animals [Noda et al., 2001]. GLP-2 also inhibits
apoptosis in the small intestine [Drucker, 2001]. Although no direct evidence of intestinal
mucosa proliferation was reported on human diabetics, small bowel hypertrophy was
described in a patient carrying a proglucagon-expressing tumor that secreted the intestinal
PGDPs [Gleeson et al., 1971].
Advanced glycation end products (AGEs) produced during the development of DM are
likely associated to the morphological changes and biomechanical remodeling of the GI tract
in DM. AGEs are generated by the sequential non-enzymatic glycation of protein amino
groups and by oxidation reaction [Singh et a., 2001]. The accumulation of AGEs in tissues
alters the structure and function of matrix proteins [Monnier et al, 1992]. Studies on DM and
ageing show that AGEs are causing cross-linking of collagen molecules responsible for
basement membrane thickening and loss of matrix elasticity [Ulrich et al., 2001; Sanchez et
al., 2000; Reddy, 2004]. The GI wall is rich in collagen [Gregersen, 2002]. Therefore,
alterations in the orientation, configuration, and content of collagen are likely to affect the
biomechanical properties of the GI tract based on remodeling studies in other disease models.
Some studies on the arteries in the diabetes have demonstrated that the non-enzymatic
glycation of collagen have a close relation with the increased arterial wall stiffness [Sims et
al., 1996; Wolffenbuttel et al., 1998]. We believe that the same way apply to the collagen in
the GI wall. AGEs may contribute to the diabetic GI morphological and biomechanical
remodeling by at least two major mechanisms. The first is receptor-independent alteration of
the extracellular matrix architecture by non-enzymatic glycation and the formation of protein
cross-links. The second mechanism is receptor-dependent and consists of modulation of
cellular functions through ligation of specific cell surface receptors [Bierhaus et al., 2005;
Stern et al., 2002; Wautier et al., 1996]. Sanchez SS and coworkers demonstrated that the
expression of small intestinal extracellular matrix proteins changed in STZ-induced diabetic
rats [Sanchez et al., 2000]. Furthermore, the different components in GI wall have different
relaxation functions [Fung, 1993]. For example, collagen has a large Young's modulus and it
relaxes less. Smooth muscle has a low modulus of elasticity and relaxes more. Purslow and
co-workers have studied the effect of collagen reorientation on creep and stress relaxation in
soft connective tissues [Purslow et al., 1998]. They demonstrated that the collagen
reorientation is not a primary source of their viscoelastic properties. The relaxation processes
within the collagen fibers or at the fiber-matrix interface may be responsible for their
viscoelastic nature. Following the changes of structural components in the GI wall during
diabetes, the viscoelastic properties of GI wall change as well, as we demonstrated that the
viscoelastic behavior of intestinal wall changed during the development of diabetes.

Therefore, the nonenzymatic glycation of the GI tissue induced by long-term hyperglycemia
seems to be an important mechanism behind the GI wall remodeling in DM.
Implication of GI Remodeling on Mechanosensory Assessments
Mechanosensation is of fundamental importance for the GI function. The

mechanosensitive nerve endings exist extensively in the GI tract where they serve a critical
role in homeostasis. Several mechanoreceptor-like structures have been identified such as
IGLE and IMA in the GI tract [Patterson et al., 2003., Phillips and Powley, 2000; Raab et al.,
2004; Swithers et al., 2002; Zagorodnyuk et al., 2001, 2003, 2005]. The mechanosensitive
afferents in the intrinsic and extrinsic pathways were described as low-, wide-dynamic- or
high-threshold tension-receptors. The GI wall structure or deformation changes in the DM
will alter the relative positions of the mechanosensitive afferents (zero setting of the
mechanosensitive afferents) [Pederse et al., 2003]. The biomechanical remodeling in the DM
such as alterations of residual strain and stress distribution and increase the wall stiffness will
alter the tension and stress distribution of the mechanosensitive afferents. As results, the
perception and motility of the GI tract will change as well. Hence, the morphological changes
and biomechanical remodeling of GI tract in the DM is likely to affect the function of
mechanosensitive afferents in the GI wall and further affect the motor and sensory function.
However, so far data are sparse on the association between the morphological and
biomechanical remodeling of GI tract and the motor-sensory dysfunction in the DM [Frokjaer
et al., 2007a, 2007a]. The multimodal stimulations have been proven accurate and reliable in
the assessment of visceral sensation in several studies [Drewes et al., 2002, 2003; Pedersen et
al., 2004]. The geometry data of GI tract can be obtained by impedance and cross-sectional
imaging such as ultrasound [Drewes et al., 2003; Frokjaer et al., 2005; Pedersen et al., 2005].
Combined with pressure recordings, biomechanical parameters such as tension, stress and
strain can be obtained and correlated to the symptoms and mechanosensory data in the DM
subjects [Frokjaer et al., 2007b]. Therefore, combined studies the GI motor-sensory
dysfunction and morphological and biomechanical remodeling in the diabetic GI tract will
improve our understanding about the pathophysiology of GI disorders in the DM patients.
FUTURE PERSPECTIVES
The pathogenesis for GI motor disorders in diabetes is not well understood. However, as
mentioned above biomechanical remodeling of the GI tract during the development of
diabetes may contribute to the GI motor disorders. Therefore, the study of the relation
between the GI motor dysfunction and morphological and biomechanical remodeling in the
diabetic GI tract may shed more light to understand the mechanism of GI dysfunction in the
diabetic patients.
Combine Study of Active and Passive Biomechanical Properties of GI Tract
According to the well known Hill‘s model in muscle mechanics and physiology, we
consider the GI tissue to consist of passive and active elements (Figure 9). The first step is to
analyze the passive elements belonging to the connective tissue because the passive elastic
properties reflect the state of the structural components determining a specific function or
malfunction. When the passive properties are known, future work can be directed to complete
the Hill‘s three-element model so that we may have an understanding of the interaction
between the parallel and contractile elements in diseases and in remodeling of the GI tissues.
In the future experiments will be designed to mimic physiological conditions both in vivo and
in vitro in normal and diabetic rats. The pressure in the lumen of the GI segment and the rate
of flow through the lumen can be controlled. At the same time the motility pattern of GI tract
can be measured. Through these combined measurements, we can get more useful
physiological data and further understand the mechanism of diabetes and the GI remodeling
and function.
Figure 9. The typical active phasic and tonic stress-strain relation curves from rat jejunal segment are
shown. The isometric length-tension diagram known from in vitro studies of muscle strips can be
obtained in intact intestinal segments. The total and passive stress increased in an exponential-like way
as function of strain. The active phasic and tonic stress reached a maximum of 2.54 and 1.46 kPa at a
strain of 0.78 and 0.76.
Investigation of the Mechanism for the GI Remodeling in Diabetes
The studies showed that diabetes alters the stress-strain relation of the GI tract in a way
consistent with a stiffer wall. However, no evidence of which structure in the wall causes the
increased stiffness is given. The GI submucosa is mainly composed of collagen. As we
postulated above alterations in the orientation, configuration, and content of collagen are
likely candidates based on remodeling studies in other disease models. In order to address this
issue, histochemical, immunohistochemical and electron microscopic work must be done to
study the contents, ultrastructure and orientation of collagen in the intestine of diabetic rats.
The analysis of advanced non-enzymatic glycation ending products (AGEs) and receptors of
AGEs expression at diabetic Gl tissues will also be subject of study.
The morphological remodeling of GI tract in the diabetes is likely to affect the zero-
setting of mechanoreceptors in the GI wall. Therefore, the identification of mechanoreceptors
will obviously be beneficial to understand the mechanisms of GI remodeling during the
diabetes. Furthermore, as mentioned in the literature, the motility disorder of GI tract is likely
closely associated with the nervous dysfunction in the GI wall. Therefore, the next step is to
investigate the enteric nerve function combined with the biomechanical properties of GI tract
in the diabetic animal model.
CONCLUSION
GI symptoms are frequent in the diabetic patients and are associated with sensory-motor
abnormalities, such as impaired perception and motility of the GI tract. The pathogenesis of
abnormal GI sensory-motor function in DM is clearly multi-factorial. The morphological
changes and biomechanical remodeling of the GI wall may compromise the GI motor
function and affect the function of the mechanosensitive afferents in the GI wall. Studies of
the relation between the GI motor-sensory dysfunction, morphological changes and
biomechanical remodeling in the diabetic GI tract may shed of more light to understand the
mechanism of GI motor-sensory dysfunction in the diabetic patients. This knowledge may
prove to be valuable in the development of new treatment strategies, such as using agents
which can break down already formed glycation end product protein-protein crosslinks.
ACKNOWLEDGMENTS
Karen Elise Jensens Foundations supported this work.
REFERENCES
Adachi T, Mori C, Sakurai K, Shihara N, Tsuda K, Yasuda K. Morphological changes and
increased sucrase and isomaltase activity in small intestines of insulin-deficient and type
2 diabetic rats. Endocr J. 2003; 50: 271-279.
Adeghate, E., P. Schattner, and E. Dunn. "An update on the etiology and epidemiology of
diabetes mellitus." Ann. N Y Acad. Sci. 2006; 1084: 1-29.
Anderson RP. Coeliac disease. Aust Fam Physician. 2005; 34: 239-242.
Anjaneyulu M, Ramarao P. Studies on gastrointestinal tract functional changes in diabetic
animals. Methods Find. Exp. Clin. Pharmacol. 2002; 24: 71-75.
Annese V, Bassotti G, Caruso N, De Cosmo S, Gabbrielli A, Modoni S, Frusciante V,
Andriulli A. Gastrointestinal motor dysfunction, symptoms, and neuropathy in
noninsulin-dependent (type 2) diabetes mellitus. J. Clin. Gastroenterol. 1999; 29: 171-

177.
Battle WM, Snape WJ Jr, Alavi A, Cohen S, Braunstein S. Colonic dysfunction in diabetes
mellitus. Gastroenterology. 1980; 79: 1217-1221.
Battle WM, Cohen JD, Snape WJ Jr. Disorders of colonic motility in patients with diabetes
mellitus. Yale J. Biol. Med. 1983; 56: 277-283.
Bhor VM, Raghuram N, Sivakami S. Oxidative damage and altered antioxidant enzyme
activities in the small intestine of streptozotocin-induced diabetic rats. Int. J. Biochem.
Cell Biol. 2004; 36: 89-97.
Bierhaus A, Humpert PM, Morcos M et al. Understanding RAGE, the receptor for advanced
glycation end products. J. Mol. Med. 2005; 83: 876-886.
Bjornsson ES, Urbanavicius V, Eliasson B, Attvall S, Smith U, Abrahamsson H. Effects of
hyperglycemia on interdigestive gastrointestinal motility in humans. Scand J.
Gastroenterol. 1994; 29: 1096-1104.
Brubaker PL, Stobie KM, Roberge JN, Lui EY, Drucker DJ. Proglucagon-derived peptides in
the neuroendocrine system. Adv. Exp. Med. Biol. 1991: 291: 143-159.
Camilleri M, Malagelada JR. Abnormal intestinal motility in diabetics with the gastroparesis
syndrome. Eur. J. Clin. Invest. 1984; 14: 420-427.
Charlton M, Ahlman B, Nair KS. The effect of insulin on human small intestinal mucosal
protein synthesis. Gastroenterology. 2000:118: 299-306.
de Boer SY, Masclee AA, Lam WF, Schipper J, Jansen JB, Lamers CB. Hyperglycemia
modulates gallbladder motility and small intestinal transit time in man. Dig. Dis. Sci.
1993; 38: 2228-2235.
Dooley CP, el Newihi HM, Zeidler A, Valenzuela JE. Abnormalities of the migrating motor
complex in diabetics with autonomic neuropathy and diarrhea. Scand J. Gastroenterol.
1988; 23: 217-223.
Drewes AM, Schipper KP, Dimcevski G, Petersen P, Andersen OK, Gregersen H, Arendt-
Nielsen L. et al. Multimodal assessment of pain in the esophagus: a new experimental
model. Am. J. Physiol. Gastrointest Liver Physiol. 2002; 283: G95-103.
Drewes AM, Gregersen H, Arendt-Nielsen L. Experimental pain in gastroenterology: a
reappraisal of human studies. Scand J Gastroenterol. 2003; 38: 1115-1130.
Drewes AM, Pedersen J, Liu W, Arendt-Nielsen L, Gregersen H. Controlled mechanical
distension of the human oesophagus: sensory and biomechanical findings. Scand. J.
Gastroenterol. 2003; 38: 27-35.
Drucker DJ. Glucagon-like peptide 2. J. Clin. Endocrinol Metab. 2001; 86: 1759-1764.
Ejskjaer NT, Bradley JL, Buxton-Thomas MS, Edmonds ME, Howard ER, Purewal T,
Thomas PK, Watkins PJ. Novel surgical treatment and gastric pathology in diabetic
gastroparesis. Diabet Med.1999; 16: 488-495.
el Salhy M. Gastrointestinal transit in nonobese diabetic mouse: an animal model of human
diabetes type 1. J. Diabetes Complications. 2001; 15: 277-284.
el Salhy M. Gastrointestinal transit in an animal model of human diabetes type 2: relationship
to gut neuroendocrine peptide contents. Ups J. Med. Sci. 2002; 107: 101-110.
Fischer KD, Dhanvantari S, Drucker DJ, Brubaker PL. Intestinal growth is associated with
elevated levels of glucagon-like peptide 2 in diabetic rats. Am. J. Physiol. 1997; 273:
E815-E820.
Folwaczny C, Hundegger K, Volger C, Sorodoc J, Kuhn M, Tatsch K, Landgraf R, Karbach

U. Measurement of transit disorders in different gastrointestinal segments of patients with
diabetes mellitus in relation to duration and severity of the disease by use of the metal-
detector test. Z. Gastroenterol. 1995; 33: 517-526.
Folwaczny C, Riepl R, Tschop M, Landgraf R. Gastrointestinal involvement in patients with
diabetes mellitus: Part I (first of two parts). Epidemiology, pathophysiology, clinical
findings. Z. Gastroenterol. 1999: 37: 803-815.
Forrest A, Parsons M. The enhanced spontaneous activity of the diabetic colon is not the
consequence of impaired inhibitory control mechanisms. Auton Autacoid. Pharmacol.
2003; 23: 149-158.
Forrest A, Huizinga JD, Wang XY, Liu LW, Parsons M. Increase in stretch-induced rhythmic
motor activity in the diabetic rat colon is associated with loss of ICC of the submuscular
plexus. Am. J. Physiol. Gastrointest Liver Physiol. 2008; 294: G315-G326.
Forster J, Damjanov I, Lin Z, Sarosiek I, Wetzel P, McCallum RW. Absence of the interstitial
cells of Cajal in patients with gastroparesis and correlation with clinical findings. J.
Gastrointest Surg 2005; 9: 102-108.
Frokjaer JB, Andersen SD, Gale J, Arendt-Nielsen L, Gregersen H, Drewes AM. An
experimental study of viscero-visceral hyperalgesia using an ultrasound-based
multimodal sensory testing approach. Pain. 2005; 119: 191-200.
Frokjaer JB, Andersen SD, Ejskaer N, Funch-Jensen P, Arendt-Nielsen L, Gregersen H,
Drewes AM. Gut sensations in diabetic autonomic neuropathy. Pain. 2007a; 131(3): 320-
329.
Frokjaer JB, Andersen SD, Ejskjaer N, Funch-Jensen P, Drewes AM, Gregersen H. Impaired
contractility and remodeling of the upper gastrointestinal tract in diabetes mellitus type-1.
World J. Gastroenterol. 2007b; 13: 4881-4890.
Fung YC. Biomechanics. Properties of living tissues. Berlin: Springer-Verlag, 1993.
Gleeson MH, Bloom SR, Polak JM, Henry K, Dowling RH. Endocrine tumour in kidney
affecting small bowel structure, motility, and absorptive function. Gut. 1971; 12: 773-
782.
Granneman JG, Stricker EM. Food intake and gastric emptying in rats with streptozotocin-
induced diabetes. Am. J. Physiol. 1984: 247: R1054-R1061.
Gregersen H. Biomechanics of the Gastrointestinal Tract. London: Springer-Verlag, 2002.
Guo C, Quobatari A, Shangguan Y, Hong S, Wiley JW. Diabetic autonomic neuropathy:
evidence for apoptosis in situ in the rat. Neurogastroenterol Motil. 2004; 16: 335-345.
He CL, Soffer EE, Ferris CD, Walsh RM, Szurszewski JH, Farrugia G. Loss of interstitial
cells of cajal and inhibitory innervation in insulin-dependent diabetes. Gastroenterology.
2001; 121: 427-434.
Holloway RH, Tippett MD, Horowitz M, Maddox AF, Moten J, Russo A. Relationship
between esophageal motility and transit in patients with type I diabetes mellitus. Am. J.
Gastroenterol. 1999; 94: 3150-3157.
Horowitz M, Maddox AF, Wishart JM, Harding PE, Chatterton BE, Shearman DJ.
Relationships between oesophageal transit and solid and liquid gastric emptying in
diabetes mellitus. Eur. J. Nucl. Med. 1991; 18: 229-234.
Horowitz M, Wishart JM, Jones KL, Hebbard GS. Gastric emptying in diabetes: an overview.
Diabet Med. 1996; 13: S16-S22.
Horowitz M, Samsom M. Gastrointestinal function in diabetes mellitus. Chichester: John

Wiley and Sons, Ltd, 2004.
Horvath VJ, Vittal H, Ordog T. Reduced insulin and IGF-I signaling, not hyperglycemia,
underlies the diabetes-associated depletion of interstitial cells of Cajal in the murine
stomach. Diabetes. 2005; 54: 1528-1533.
Iber FL, Parveen S, Vandrunen M, Sood KB, Reza F, Serlovsky R, Reddy S. Relation of
symptoms to impaired stomach, small bowel, and colon motility in long-standing
diabetes. Dig. Dis. Sci. 1993; 38: 45-50.
Iida M, Ikeda M, Kishimoto M, Tsujino T, Kaneto H, Matsuhisa M, Kajimoto Y, Watarai T,
Yamasaki Y, Hori M. Evaluation of gut motility in type II diabetes by the radiopaque
marker method. J. Gastroenterol Hepatol. 2000; 15: 381-385.
Iyer SK, Chandrasekhara KL, Sutton A. Diffuse muscular hypertrophy of esophagus. Am J
Med. 1986; 80: 849-852.
James AN, Ryan JP, Crowell MD, Parkman HP. Regional gastric contractility alterations in a
diabetic gastroparesis mouse model: effects of cholinergic and serotoninergic stimulation.
Am. J. Physiol. Gastrointest Liver Physiol. 2004; 287: G612-G619.
Jervis EL, Levin RJ. Anatomic adaptation of the alimentary tract of the rat to the hyperphagia
of chronic alloxan-diabetes. Nature. 1966; 210: 391-393.
Jones KL, Horowitz M, Wishart MJ, Maddox AF, Harding PE, Chatterton BE . Relationships
between gastric emptying, intragastric meal distribution and blood glucose concentrations
in diabetes mellitus. J. Nucl. Med 1995; 36: 2220-2228.
Jorgensen F, Boesen F, Andersen EB, Hesse B. Oesophageal transit in patients with
autonomic dysfunction. The effect of treatment with fludrocortisone. Clin. Physiol. 1991;
11: 83-92.
Jorgensen CS, Ahrensberg JM, Gregersen H, Flyvberg A. Tension-strain relations and
morphometry of rat small intestine in experimental diabetes. Dig. Dis. Sci. 2001: 46: 960-
967.
Jung HK, Kim DY, Moon IH, Hong YS. Colonic transit time in diabetic patients--comparison
with healthy subjects and the effect of autonomic neuropathy. Yonsei Med. J. 2003;
44(2): 265-272.
Kamath MV, Tougas G, Fitzpatrick D et al. Assessment of the visceral afferent and
autonomic pathways in response to esophageal stimulation in control subjects and in
patients with diabetes. Clin. Invest. Med. 1998: 21: 100-113.
Kamath MV, May A, Hollerbach S et al. Effects of esophageal stimulation in patients with
functional disorders of the gastrointestinal tract. Crit Rev Biomed Eng. 2000: 28: 87-93.
Keshavarzian A, Iber FL, Dangleis MD, Cornish R. Intestinal-transit and lactose intolerance
in chronic alcoholics. Am. J. Clin. Nutr. 1986; 44: 70-76.
Ko GT, Chan WB, Chan JC, Tsang LW, Cockram CS. Gastrointestinal symptoms in Chinese
patients with Type 2 diabetes mellitus. Diabet Med. 1999; 16: 670-674.
Korolkiewicz R, Rekowski P, Szyk A, Kato S, Yasuhiro T, Kubomi M, Tashima K, Takeuchi
K. Effects of diabetes mellitus on the contractile activity of carbachol and galanin in
isolated gastric fundus strips of rats. Pharmacology. 1998; 57: 65-78.
Liao D, Zhao J, Gregersen H. Three-dimensional geometry analysis of the stomach in type II
diabetic GK rats. Diabetes Res. Clin. Pract. 2006; 71: 1-13.
Lincoln J, Bokor JT, Crowe R, Griffith SG, Haven AJ, Burnstock G. Myenteric plexus in
streptozotocin-treated rats. Neurochemical and histochemical evidence for diabetic
neuropathy in the gut. Gastroenterology. 1984; 86 654-661.
Liu J, Qiao X, Micci MA, Pasricha PJ, Chen JD. Improvement of gastric motility with gastric
electrical stimulation in STZ-induced diabetic rats. Digestion. 2004; 70: 159-166.
Long QL, Fang DC, Shi HT, Luo YH. Gastro-electric dysrhythm and lack of gastric
interstitial cells of cajal. World J. Gastroenterol. 2004; 10: 1227-1230.
Longstreth GF. Diabetic thoracic polyradiculopathy: ten patients with abdominal pain. Am. J.
Gastroenterol. 1997; 92: 502-505.
Maleki D, Locke GR, Camilleri M 3rd, Zinsmeister AR, Yawn BP, Leibson C, Melton LJ
3rd. Gastrointestinal tract symptoms among persons with diabetes mellitus in the
community. Arch. Intern. Med. 2000; 160: 2808-2816.
Mayhew TM, Carson FL, Sharma AK. Small intestinal morphology in experimental diabetic
rats: a stereological study on the effects of an aldose reductase inhibitor (ponalrestat)
given with or without conventional insulin therapy. Diabetologia. 1989: 32: 649-654.
Mehta N, Veliath S, Thombre DP. Effect of experimental diabetes and vagotomy on gastric
emptying in rats. Indian J. Physiol. Pharmacol. 2002; 46: 441-448.
Miller DL, Hanson W, Schedl HP, Osborne JW. Proliferation rate and transit time of mucosal
cells in small intestine of the diabetic rat. Gastroenterology. 1977; 73: 1326-1332.
Monnier VM, Sell DR, Nagaraj RH, Miyata S, Grandhee S, Odetti P, Ibrahim SA. Maillard
reaction-mediated molecular damage to extracellular matrix and other tissue proteins in
diabetes, aging, and uremia. Diabetes. 1992; 41 Suppl 2: 36-41.
Nair, M. "Diabetes mellitus, part 1: physiology and complications." Br. J. Nurs. 2007; 16:
184-188.
Nakahara M, Isozaki K, Hirota S, Vanderwinden JM, Takakura R, Kinoshita K, Miyagawa J,
Chen H, Miyazaki Y, Kiyohara T, Shinomura Y, Matsuzawa Y. Deficiency of KIT-
positive cells in the colon of patients with diabetes mellitus. J. Gastroenterol. Hepatol.
2002; 17: 666-670.
Nguyen HN, Silny J, Wuller S, Marschall HU, Rau G, Matern S. Abnormal postprandial
duodenal chyme transport in patients with long standing insulin dependent diabetes
mellitus. Gut. 1997; 41: 624-631.
Noda T, Iwakiri R, Fujimoto K, Yoshida T, Utsumi H, Sakata H, Hisatomi A, Aw TY.
Suppression of apoptosis is responsible for increased thickness of intestinal mucosa in
streptozotocin-induced diabetic rats. Metabolism. 2001; 50: 259-264.
Nowak TV, Roza AM, Weisbruch JP, Brosnan MR. Accelerated gastric emptying in diabetic
rodents: effect of insulin treatment and pancreas transplantation. J. Lab. Clin. Med. 1994;
123: 110-116.
Ogata M, Iizuka Y, Murata R, Hikichi N. Effect of streptozotocin-induced diabetes on
cyclosporin A disposition in rats. Biol. Pharm. Bull. 1996; 19: 1586-1590.
Ogata M, Uchimura T, Iizuka Y Murata R, Suzuki S, Toyota T, Hikichi N. Effect of non-
insulin dependent diabetes on cyclosporin A disposition in Goto-Kakizaki (GK) rats.
Biol. Pharm. Bull. 1997; 20: 1026-1029.
Patterson LM, Zheng H, Ward SM, Berthoud HR. Vanilloid receptor (VR1) expression in
vagal afferent neurons innervating the gastrointestinal tract. Cell Tissue Res. 2003; 311:
277-287.
Pedersen J, Gao C, Egekvist H, Bjerring P, Arendt-Nielsen L, Gregersen H, Drewes AM.

Pain and biomechanical responses to distention of the duodenum in patients with
systemic sclerosis. Gastroenterology. 2003; 124: 1230-1239.
Pedersen J, Reddy H, Funch-Jensen P, Arendt-Nielsen L, Gregersen H, Drewes AM. Cold
and heat pain assessment of the human oesophagus after experimental sensitisation with
acid. Pain. 2004; 110: 393-399.
Pedersen J, Drewes AM, Gregersen H. New analysis for the study of the muscle function in
the human oesophagus. Neurogastroenterol. Motil. 2005; 17: 767-772.
Phillips RJ, Powley TL. Tension and stretch receptors in gastrointestinal smooth muscle: re-
evaluating vagal mechanoreceptor electrophysiology. Brain Res. Brain Res. Rev. 2000;
34: 1-26.
Purslow PP, Wess TJ, Hukins DW. Collagen orientation and molecular spacing during creep
and stress-relaxation in soft connective tissues. J. Exp. Biol. 1998: 201: 135-142.
Raab M, Neuhuber WL. Intraganglionic laminar endings and their relationships with neuronal
and glial structures of myenteric ganglia in the esophagus of rat and mouse. Histochem.
Cell Biol. 2004; 122: 445-459.
Rathmann W, Enck P, Frieling T, Gries FA. Visceral afferent neuropathy in diabetic
gastroparesis. Diabetes Care. 1991; 14: 1086-1089.
Rauma J, Spångeus A, El-Salhy M. Ghrelin cell density in the gastrointestinal tracts of animal
models of human diabetes. Histol. Histopathol. 2006; 21: 1-5.
Rayner CK, Verhagen MA, Hebbard GS, DiMatteo AC, Doran SM, Horowitz M. Proximal
gastric compliance and perception of distension in type 1 diabetes mellitus: effects of
hyperglycemia. Am. J. Gastroenterol. 2000; 95: 1175-1183.
Reddy GK. AGE-related cross-linking of collagen is associated with aortic wall matrix
stiffness in the pathogenesis of drug-induced diabetes in rats. Microvasc. Res. 2004; 68:
132-142.
Ricci JA, Siddique R, Stewart WF, Sandler RS, Sloan S, Farup CE. Upper gastrointestinal
symptoms in a U.S. national sample of adults with diabetes. Scand. J. Gastroenterol.
2000; 35: 152-159.
Samsom M, Jebbink RJ, Akkermans LM, Berge-Henegouwen GP, Smout AJ. Abnormalities
of antroduodenal motility in type I diabetes. Diabetes Care. 1996; 19: 21-27.
Samsom M, Roelofs JM, Akkermans LM, Berge Henegouwen GP, Smout AJ. Proximal
gastric motor activity in response to a liquid meal in type I diabetes mellitus with
autonomic neuropathy. Dig. Dis. Sci. 1998; 43: 491-496.
Samsom M, Verhagen MAMT. Intestinal function. In: Horowitz M, Samsom M, editor.
Gastrointestinal function in diabetes mellitus. Chichester: John Wiley and Sons, Ltd,
2004; 177-218.
Sanchez SS, Genta SB, Aybar MJ, Honore SM, Villecco EI, Sanchez Riera AN. Changes in
the expression of small intestine extracellular matrix proteins in streptozotocin-induced
diabetic rats. Cell Biol. Int. 2000; 24: 881-888.
Schvarcz E, Palmer M, Ingberg CM, Aman J , Berne C. Increased prevalence of upper
gastrointestinal symptoms in long-term type 1 diabetes mellitus. Diabet. Med. 1996; 13:
478-481.
Secondulfo M, Iafusco D, Carratu R, deMagistris L, Sapone A, Generoso M, Mezzogiomo A,
Sasso FC, Carteni M, De Rosa R, Prisco F, Esposito V. Ultrastructural mucosal
alterations and increased intestinal permeability in non-celiac, type I diabetic patients.

Dig. Liver Dis. 2004; 36: 35-45.
Sha H, Zhao JB, Zhang ZY, Zhou SP, Tong XL, Zhuang FY, Gregersen H. Effect of Kaiyu
Qingwei Jianji on the morphometry and residual strain distribution of small intestine in
experimental diabetic rats. World J. Gastroenterol. 2006; 12: 7149-7154.
Sims MA, Hasler WL, Chey WD, Kim MS, Owyang C. Hyperglycemia inhibits
mechanoreceptor-mediated gastrocolonic responses and colonic peristaltic reflexes in
healthy humans. Gastroenterology. 1995; 108: 350-359.
Sims TJ, Rasmussen LM, Oxlund H, Bailey AJ. The role of glycation cross-links in diabetic
vascular stiffening. Diabetologia. 1996: 39: 946-951.
Singh R, Barden A, Mori T, Beilin L. Advanced glycation end-products: a review.
Diabetologia. 2001; 44: 129-146.
Smith DS, Ferris CD. Current concepts in diabetic gastroparesis. Drugs. 2003; 63: 1339-
1358.
Smout AJPM. Oesophageal function. In: Horowitz M, Samsom M, editor. Gastrointestinal
function in diabetes mellitus. Chichester: John Wiley and Sons, Ltd, 2004; 97-116.
Spangeus A, El-Salhy M. Large intestinal endocrine cells in non-obese diabetic mice. J.
Diabetes Complications. 1998; 12: 321-327.
Spangeus A, el Salhy M, Suhr O, Eriksson J, Lithner F. Prevalence of gastrointestinal
symptoms in young and middle-aged diabetic patients. Scand. J. Gastroenterol. 1999; 34:
1196-1202.
Spangeus A, Kand M, El-Salhy M. Gastrointestinal endocrine cells in an animal model for
human type 2 diabetes. Dig. Dis. Sci. 1999; 44: 979-985.
Stern DM, Yan SD, Yan SF, Schmidt AM. Receptor for advanced glycation endproducts
(RAGE) and the complications of diabetes. Ageing Res. Rev. 2002; 1: 1-15.
Sun WM, Read NW. Anorectal function. In: Horowitz M, Samsom M. editor.
Gastrointestinal function in diabetes mellitus. Chichester: John Wiley and Sons, Ltd,
2004; 219-246.
Swithers SE, Baronowsky E, Powley TL. Vagal intraganglionic laminar endings and
intramuscular arrays mature at different rates in pre-weanling rat stomach. Auton
Neurosci. 2002; 102: 13-19.
Tahara T, Yamamoto T. Morphological changes of the villous microvascular architecture and
intestinal growth in rats with streptozotocin-induced diabetes. Virchows Arch. A. Pathol.
Anat. Histopathol. 1988: 413: 151-158.
Talley NJ, Young L, Bytzer P, Hammer J, Leemon M, Jones M, Horowitz M. Impact of
chronic gastrointestinal symptoms in diabetes mellitus on health-related quality of life.
Am. J. Gastroenterol. 2001; 96: 71-76.
Talley NJ, Howell S, Jones MP, Horowitz M. Predictors of turnover of lower gastrointestinal
symptoms in diabetes mellitus. Am. J. Gastroenterol. 2002; 97: 3087-3094.
Tormo MA, Martinez IM, Romero dT, Gil-Exojo I, Campillo JE. Morphological and
enzymatic changes of the small intestine in an n0-STZ diabetes rat model. Exp. Clin.
Endocrinol. Diabetes. 2002; 110: 119-123.
Tsai SC, Kao CH, Pan DY, ChangLai SP, Wang SJ. Effects of oral erythromycin on
esophageal motility in patients with noninsulin-dependent diabetes mellitus. Gaoxiong Yi
Xue Ke Xue Za Zhi. 1995; 11: 430-435.
Ulrich P, Cerami A. Protein glycation, diabetes, and aging. Recent Prog. Horm. Res. 2001;
56: 1-21.
Umpierrez G, Freire AX. Abdominal pain in patients with hyperglycemic crises. J. Crit.
Care. 2002; 17: 63-67.
Unal A, Guven K, Yurci A, Torun E, Gursoy S, Baskol M, Ozturk F, Arsav V. Is increased
colon subepithelial collagen layer thickness in diabetic patients related to collagenous
colitis? An immunohistochemical study. Pathol. Res. Pract. 2008; 204: 537-544.
Undeland KA, Hausken T, Gilja OH, Aanderud S, Berstad A. Gastric meal accommodation
and symptoms in diabetes. A placebo-controlled study of glyceryl trinitrate. Eur J
Gastroenterol Hepatol. 1998; 10: 677-681.
Vahl TP, D'Alessio DA. Gut peptides in the treatment of diabetes mellitus. Expert. Opin.
Investig Drugs. 2004; 13:177-188.
Verne GN, Sninsky CA. Diabetes and the gastrointestinal tract. Gastroenterol Clin. North
Am. 1998: 27: 861-874.
Watanabe T, Asanuma A, Tanaka M, Akiba T, Koga T. [Morphological study on the gastric
mucosa in diabetes mellitus rats induced by streptozotocin]. Exp. Anim. 1995; 43: 693-
696.
Wautier JL, Zoukourian C, Chappey O, Wautier MP, Guillausseau PJ, Cao R, Hori O, Stern
D, Schmidt AM. Receptor-mediated endothelial cell dysfunction in diabetic
vasculopathy. Soluble receptor for advanced glycation end products blocks
hyperpermeability in diabetic rats. J. Clin. Invest. 1996; 97: 238-243.
Wolffenbuttel BH, Boulanger CM, Crijns FR et al. Breakers of advanced glycation end
products restore large artery properties in experimental diabetes. Proc. Natl. Acad. Sci. U
S A. 1998: 95: 4630-4634.
Yang J, Zhao J, Zeng Y, Gregersen H. Biomechanical properties of the rat oesophagus in
experimental type-1 diabetes. Neurogastroenterol Motil. 2004; 16: 195-203.
Yang J, Zhao J, Liao D, Gregersen H. Biomechanical properties of the layered oesophagus
and its remodelling in experimental type-1 diabetes. J Biomech. 2006; 39: 894-904.
Yoneda S, Kadowaki M, Kuramoto H, Fukui H, Takaki M. nhanced colonic peristalsis by
impairment of nitrergic enteric neurons in spontaneously diabetic rats. Auton Neurosci.
2001; 92: 65-71.
Young AA, Gedulin B, Vine W, Percy A, Rink TJ. Gastric emptying is accelerated in diabetic
BB rats and is slowed by subcutaneous injections of amylin. Diabetologia. 1995; 38:
642-648.
Zagorodnyuk VP, Chen BN, Brookes SJ. Intraganglionic laminar endings are mechano-
transduction sites of vagal tension receptors in the guinea-pig stomach. J. Physiol. 2001;
534: 255-268.
Zagorodnyuk VP, Chen BN, Costa M, Brookes SJ. Mechanotransduction by intraganglionic
laminar endings of vagal tension receptors in the guinea-pig oesophagus. J. Physiol.
2003; 553: 575-587.
Zagorodnyuk VP, Lynn P, Costa M, Brookes SJ. Mechanisms of mechanotransduction by
specialized low-threshold mechanoreceptors in the guinea pig rectum. Am. J. Physiol.
Gastrointest Liver Physiol. 2005; 289: G397-G406.
Zeng YJ, Yang J, Zhao JB, Liao DH, Zhang EP, Gregersen H, Xu XH, Xu H, Xu CQ.
Morphologic and biomechanical changes of rat oesophagus in experimental diabetes.
World J. Gastroenterol. 20041; 10: 2519-2523.
Zhao J, Sha H, Zhou S, Tong X, Zhuang FY, Gregersen H. Remodelling of zero-stress state
of small intestine in streptozotocin-induced diabetic rats. Effect of gliclazide. Dig. Liver
Dis. 2002; 34: 707-716.
Zhao J, Liao D, Yang J, Gregersen H. Viscoelastic behavior of small intestine in
streptozotocin-induced diabetic rats. Dig. Dis. Sci. 2003a; 48: 2271-2277.
Zhao J, Yang J, Gregersen H. Biomechanical and morphometric intestinal remodelling during
experimental diabetes in rats. Diabetologia. 2003b; 46: 1688-1697.
Zhao J, Frøkjaer JB, Drewes AM, Ejskjaer N. Upper gastrointestinal sensory-motor
dysfunction in diabetes mellitus. World J. Gastroenterol. 2006; 12: 2846-2857.
Zhao J, Liao D, Gregersen H. Biomechanical and histomorphometric esophageal remodeling
in type 2 diabetic GK rats. J. Diabetes Complications. 2007; 21: 34-40.
Zhao J, Nakaguchi T, Gregersen H. Biomechanical and Histomorphometric Colon
Remodelling in STZ-Induced Diabetic Rats. Dig. Dis. Sci. 2009; 54: 1636-1642
Zoubi SA, Mayhew TM, Sparrow RA. The small intestine in experimental diabetes: cellular
adaptation in crypts and villi at different longitudinal sites. Virchows Arch. 1995: 426:
501-507.
Chapter 5
BIOMECHANICS OF THE GASTROINTESTINAL TRACT

IN HEALTH AND DISEASE
Jingbo Zhao1, Donghua Liao1 and Hans Gregersen1,2

1
Mech-Sense, Aalborg Hospital Science and Innovation Centre (AHSIC),
9000 Aalborg, Denmark
2
La Jolla Bioengineering Institute, La Jolla, California, USA
ABSTRACT
The gastrointestinal (GI) tract is functionally subjected to dimensional changes.
Hence, biomechanical properties such as the stress-strain relationships are of particularly
importance. These properties vary along the normal GI tract and remodel in response to
growth, aging and disease. The biomechanical properties are crucial for GI motor
function because peristaltic motion that propels the food through the GI tract is a result of
interaction of the passive and active tissue forces and the hydrodynamic forces in the
food bolus and remodeling of the mechanical properties reflects the changes in the tissue
structure that determine a specific motor dysfunction. Therefore, biomechanical data on
the GI wall are important to understand the pathogenesis to the GI motor-sensory
function and dysfunction. Moreover, biomechanical studies of the GI tract pave the way
for further mathematical and computational modelling. Biomechanical studies of the GI
tract will advance our understanding of GI physiological function, diseases such as
dyspepsia and visceral pain, and GI dysfunction due to systemic diseases. Furthermore,
integrated GI simulation models will be beneficial for medical education and for
evaluation of the efficacy and safety of new drugs on GI function.
Keywords: Gastrointestinal tract, biomechanics, remodeling, diseases, modeling.

Telephone: +45 99326907, E-mail: jz@rn.dk, jzhao@hst.aau.dk
164 Jingbo Zhao, Donghua Liao and Hans Gregersen
INTRODUCTION
The major function of the gastrointestinal (GI) tract is digestion facilitated by motility,
secretion and absorption. The movement of luminal contents is facilitated by changes in the
geometry of the GI wall and lumen due to the motility. Therefore, the biomechanical
properties of the GI tract are important for its function [Gregersen, 2002a]. The mechanical
properties of the GI tract are important determinants of flow, the stress distribution in the
wall, and the remodeling process of the wall under physical environment change [Gregersen
and Kassab 1996; Gregersen 2002a]. In order to arrive at useful approximations, it is
important to understand the basic relation between stress and strain. In addition to
determining the distensibility of the GI tract, a more general relationship between stress and
strain is desirable for a more thorough understanding of the mechanical properties of the GI
tract. The stress-strain relationship must ideally be determined in reference to the zero-stress
state. Furthermore, the biomechanical properties will remodel due to normal growth
[Gregersen et al., 2004; Christensen et al., 1992a], aging [Christensen et al., 1992b], surgical
invention [Storckholm et al., 2007], systemic diseases [Zhao et al., 2006], local disorders
[Yang et al., 2009] and etc. Understanding the biomechanical properties of GI tract and its
remodeling aids the understanding of the physiology and pathophysiology of GI tract.
Medical imaging methods such as ultrasonography, magnetic resonance imaging (MRI)
and endoscopic ultrasound (EUS) are well known stand-alone clinical methods that can
disclose structural and functional abnormalities of the GI tract [Berstad et al., 1996, 1998;
Frokjaer et al., 2005, 2007a; Gilja et al., 2006, 2007]. However, a thorough study of the GI
tract requires understanding of the interactions between cells, tissues and GI organs in health
and disease. This depends on knowledge, not only of cellular ionic current mechanisms and
signal transduction pathways but also of large scale GI tissue structures and the distribution of
the nervous network. A unique way of coping with this complexity is mathematical and
computational modelling, providing a computational framework for the multilevel modelling
and simulation of the human GI anatomy and physiology [Clapworthy et al., 2007; Hunter et
al., 2002, 2005]. In GI tract studies, this approach is not common, mainly because we still
lack models that could emulate the behaviour of the human body. Nevertheless, exploration
of the GI tract has dramatically improved by the introduction of cross sectional imaging
modalities such as Computed Tomography (CT) and MRI that have revolutionised the way in
which many conditions are diagnosed and treated. The ability to examine in detail structures
inside the GI tract, without resorting to surgery, has allowed clinicians to diagnose problems
and plan corrective procedures with a minimum of risk to the patient [Gregersen, 2006; Liao
et al., 2008]. In order to continue this exploration, it will be necessary to complement the
traditional approach with an integrative approach that combines observations, theories and
predictions across the temporal and dimensional scales, across scientific disciplines, and
across the anatomical subsystems, all of which reflect the rather artificial divisions described.
The aim of this review is to describe the normal biomechanical properties, the
biomechanical remodelling due to various causes, and the currently developed biomechanical
modelling works in the GI tract on humans and animals, that can be further used to integrate
the physiological, anatomical and medical knowledge of the GI system.
Biomechanics of the Gastrointestinal Tract in Health and Disease 165
BIOMECHANICAL PROPERTIES OF NORMAL GI TISSUES

The GI tract consists of a series of organs which resemble one another in constitution,
being variously arranged as cylinders, spheroids, or intermediate forms. The main functions
of the GI tract are transport and digestion of food. The different segments show a big
variation in morphology and muscle mechanical properties. The esophagus mainly serves to
quickly transport the food bolus from the mouth to the stomach where the food in the stomach
is stored for some time and simultaneously being broken down into smaller components. The
small intestine is the longest section of the digestive tract and the chyme is mixed, absorbed
and transported to large intestine. Finally the large intestine absorbs the remaining water and
electrolytes from indigestible food matter, stores food remains that were not digested and
eliminate solid waste (feces) from the body. The movement of luminal contents in the GI tract
is facilitated by changes in the geometry of the GI wall and lumen due to the motility. It is
well known that a bolus propelled in front of a peristaltic contraction distends the gut wall
[Siegle et al., 1987, 1989, 1990]. Therefore, in order to understand the morphological and
biomechanical GI remodeling due to normal growth, aging and different diseases, it is
important to know the biomechanical properties of the normal GI tract.
Biomechanical Properties of Normal Esophagus
The mechanical properties of the esophagus are important for its function because the
esophagus is subjected to changes in wall stress and strains caused by the passage of boli and
the action of peristalsis.
Using impedance planimetry (IP) the cross-sectional area and pressure changes of the
esophagus can be measured at the same time and the esophageal wall tension and stiffness
can be obtained in vivo or in vitro. Using this novel technique it was demonstrated that the
circumferential wall tension and wall stiffness of the esophageal wall increased exponentially
as function of pressure [Orvar et al., 1993; Rao et al., 1995; Patel et al., 1998M; Vanags et al.,
2003]. The tension distribution were non-uniform along the esophagus [Patel et al.,1998;
Vanags et al., 2003] and the LES and the proximal esophagus had greater wall tension and
were less deformable than the mid- or distal esophagus [Patel et al., 1998]. Combining with
ultrasound the wall thickness of the esophagus also can be obtained and the stress-strain
relationship of the esophageal wall can be computed [Assentoft et al., 2000; Frokjaer et al.,
2006a., Zhao et al., 2007a]. Using IP combining the ultrasound, it was demonstrated that the
esophageal wall stress-strain curves are also exponential [Zhao et al., 2007a] meaning that the
tissue is soft at physiologic pressures and stiffer in the supraphysiologic pressure range. This
mechanism seems to prevent overstretch of the esophageal wall when luminal loading
becomes supraphysiologic. Furthermore it was also shown that the stress-strain distribution is
non-uniform across the esophageal wall, i.e., the stiffness increased throughout the wall and
was highest at the outer surface [Frokjaer et al., 2006a]. However, Takeda et al [2002, 2003]
showed that the stress-strain relationship of the human esophagus in vivo was linear during
both isovolumic and isobaric distension, before as well as after injection of the antimuscarinic
drug atropine. The stress-strain relationship of the active component (muscle contraction) was
different during isovolumic and isobaric distensions but the passive components were similar.
Furthermore, the method of esophageal distension significant influences on the active but not
on the passive biomechanical properties due to a strain-rate effect. Investigating the degree of
stretch and tension to initiate motor and sensory responses of the esophagus showed that the
stretch rather than tension appears to be the major factor influencing sensory responses to
esophageal distension [Barlow et al, 2002; Drewes et al, 2003a].
Esophageal function is usually evaluated by means of manometry which is a proxy of the
force in the radial direction. However, force measurements in the axial direction will provide
a better measure of esophageal transport function. Gravesen et al [2008] recently developed a
probe based on electrical impedance measurements to quantify the axial force generated by
esophageal contractions, i.e. probe elongation was associated with the axial force. The probe
showed good reproducibility and the dispersion was <0.04. The swallowed bolus size
correlated with the axial force measurements (P = 0.038) but not with manometric
measurements. Therefore, the new technique measuring axial force in the esophagus may in
the future provide valuable information about esophageal function.
Experimental and diagnostic procedures like distension of a balloon catheter, bougie, and
esophagogastroduodenoscopy can induce shear deformation in the esophageal wall. Hence,
knowing the shear modulus of the esophagus is important. Yang et al [2004a] has studied the
shear modulus of elasticity of the esophagus in the rat. Results were obtained using a triaxial
instrument to perform simultaneous torsion, inflation, and longitudinal stretching tests (Figure
1). The shear modulus varied with the inflation pressure and the longitudinal stretch ratio. The
mechanical constant of the esophagus showed that the esophageal wall was anisotropic with
different stiffness in the circumferential, longitudinal, and the shear directions. The stiffness
in the longitudinal direction was higher than in the circumferential direction.
To understand esophageal function, it is necessary to know the stress-strain relationship
for the different layers, because the esophageal wall is composed of four layers. Studies in
layered wall of the esophagus showed that the stress distribution in the different layers were
nonlinear and anisotropic [Lu and Gregersen, 2001; Gregersen et al, 2008; Fan et al, 2004;
Liao et al, 2003a, 2007; Zhao et al, 2007b; Yang et al, 2007a]. The submucosa-mucosa layer
has the highest stiffness [Fan et al., 2004; Liao et al.,2003a]. The residual strain differs
Figure1. Torsion machine setup. 1: Linear stage, 2: Torque transducer, 3: Organ bath, 4: Specimen, 5:
Force transducer, 6: Motor for rotation, 7: Pressure transducer, 8: Infusion channel, 9: Motor for linear
stage, 10: Rails for linear stage, 11: CCD camera, 12: Plastic rod.
Figure 2. Residual strain distribution in the intact wall and layers of pig oesophagus. Residual strain
distribution of the intact wall referenced to the zero-stress state of the intact wall Residual strain
distribution of layered wall in intact wall referenced to the zero-stress state of layered sectors Residual
strain distribution of separated layers referenced to the zero-stress state of layered sectors. Referenced
to the zero-stress state of the intact sectors, the inner residual strain of intact rings was -0.13 ± 0.04 and
outer residual strain was 0.31 ± 0.03. Referenced to the ―true‖ zero-stress state of separated three
layered sectors, the inner residual strain of intact rings was -0.22 ± 0.02 and outer residual strain was
0.07 ± 0.02. The residual strain distribution of the layers in intact rings referenced to the ―true‖ zero-
stress state is also shown. The inner surface residual strain was negative at mucosa-submucosa and
inner muscle layers and was positive at outer muscle layer, whereas the outer surface residual strain was
negative at the mucosa-submucosa layer and positive at the inner and outer muscle layers. Referenced
to the ―true‖ zero-stress state, the inner residual strain was negative and outer residual strain was
positive for the separated layered rings.
between the layers, and the residual strain distribution was more uniform after the layers were
separated (Figure 2) [Zhao et al, 2007b]. Furthermore, experimental studies have
demonstrated esophageal mucosal folds in the no-load state. This indicates that mucosal
buckling must be considered in the analysis of the mechanical reference state since the
material stiffness drops dramatically after tissue collapse [Liao et al, 2007; Yang et al,
2007a].
The esophageal wall is composed of different material components that contribute to the
esophageal biomechanical properties. Fan et al [2005] studied the effect of collagenase and
elastase on the morphological and biomechanical properties in the no-load and zero-stress
states in the rat esophagus. It was demonstrated that the collagenase and elastase caused the
opening angle and the residual strain in the separated mucosa-submucosa layer to decrease.
The opening angle of the separated mucosa-submucosa layer depended to some extent on the
fraction of collagen and elastin. This indicates that collagen and elastin are important for
passive biomechanical properties of the esophagus.
Biomechanical Properties of Normal Stomach
The stomach is to a larger degree than other parts of the GI tract functionally subjected to
dimensional change [Schulze-Delrieu K et al., 1998]. However, the stomach has more
complex geometry, indicating a more complex mechanical behavior. The stomach has
traditionally been studied using in vivo pressure-volume and in vitro length-tension measures
[Haffner and Stadaas, 1972; Milenov and Golenhofen, 1982; Schulze-Delrieu and Shirazi,
1987]. In a number of studies in animals and humans, the barostat was used to measure
pressure-volume relations for evaluation of tone, compliance and tension in the gastric fundus
[Azpiroz and Malagelada, 1985, 1987; Whitehead et al., 1997]. The main assumptions made
in such studies were that the fundus is geometrically a sphere with uniform tension. In other
words it is assumed that the mechanical properties in all directions are the same. However,
the geometry of the fundus of the stomach is much more complex than the spheroidal
geometry. Investigations of gastric tone must take into account the complex geometry and the
anisotropic biomechanical properties of the gastric wall.
In 1975, it was demonstrated that the properties of longitudinal and circumferential
muscle of guinea-pig stomach were different [Kuriyama et al, 1975]. This indicated that the
topical differences of the motility in the stomach may be due not only to the activity of
nervous elements, but also to differences in the properties of the muscle fibers themselves.
Egorov et al [2002] showed that the values of maximal stress and destructive strain of
stomach strips were direction-dependent. In order to investigate the anisotropic properties of
the stomach wall, Zhao et al [2005] studied uniaxial stress-strain properties of gastric strips
obtained from rats and rabbits. It was demonstrated that the biomechanical properties of the
gastric strips are location-dependent, direction-dependent and species-dependent. The
assumption in physiological pressure-volume studies that the stomach is a sphere with
uniform tension is not valid. More recently, Zhao et al [2008a] confirmed the anisotropic
biomechanical properties in pig stomach wall. Furthermore, they made an in-depth
biomechanical test on the layered wall of the stomach in different regions. It was
demonstrated that the uni-axial stress-strain curves of pig stomach were location-, direction-
and layer-dependent (Figure 3). The stiffer wall in the corpus is likely due to its thicker
mucosa, i.e., the stiffness of the mucosa-submucosa layer seems can explain the intact wall
stiffness. Since the structure and function of the pig stomach are somewhat similar to the
human stomach, the data obtained from this study can be extended to humans.
Detailed biomechanical mapping of the stomach will likely help us to understand
physiological functions of the different parts of the human stomach, such as gastric
accommodation and mechanosensation. A better understanding of gastric accommodation and
gastric perception requires knowledge of regional gastric geometry and local gastric tension
throughout the stomach. Based on medical imaging data, Liao et al [2005] described the
three-dimensional (3D) rat stomach geometry and tension distribution and showed that the
surface principal radii of curvatures and tension distribution in the stomach were non-
homogeneous.
Figure 3. Gastric strips presented as layer-dependent properties (top), Location- dependent properties
(middle) and Direction-dependent properties (bottom). The Top figure shows circumferential uniaxial
stress-strain relationships of gastric strips in fundus. The stress-strain curve for the mucosa-submucosa
strips was located to the left whereas the curve of the muscle strips was located to the right. It indicated
that the mucosa strips were the stiffest and the muscle strips were the softest. The Middle figure shows
longitudinal uniaxial stress-strain relationships of the intact gastric strips. The curves of the corpus
strips were located to the left and the curves of the fundus strips were located to the right. It indicated
that the corpus strips were the stiffest and the fundus strips were the softest. The Bottom figure shows
uniaxial stress-strain relationships of the intact gastric strips in different location. The longitudinal
stress-strain curves were located to the left, indicating that those strips were the stiffest, whereas the
curves for the circumferential strips were located to the right, indicating they were softest. Notes: F,
fundus; C, corpus; A, antrum; circ, circumferential direction; long, longitudinal direction.
Later, Liao et al [2004a] analyzed the surface geometry of the human stomach using real-
time 3-D ultrasonography in vivo to describe the three-dimensional (3-D) geometry of the
gastric antrum, gastric fundus and the whole stomach. The principal curvatures spatial
distributions were non-homogeneous in the gastric antrum, gastric fundus and the whole
stomach due to their complex geometry. The maximum longitudinal principal curvature in the
antrum, fundus and total stomach were different. As consequence providing an average
tension for the stomach does not reflect the large variation in tension throughout the stomach
wall.
Mechanical properties in the human gastric antrum were studied by Gregersen et al
[2002b] using B-mode ultrasonography and antral distension. The strain was positive in the
circumferential direction, negative in the radial direction, and with no deformation in the
longitudinal direction. Later, Gregersen et al [2006b] studied mechanosensory properties in
the human gastric antrum using the same method. The bag distensions elicited contractions in
the antrum and sensory responses below the pain threshold. Butylscopolamine abolished the
contractions and significantly reduced the sensory response. The muscle length-tension
diagram and typical preload-afterload curves as modem the Frank-Starling cardiac law can be
obtained in the human gastric antrum. The sensory responses were most closely associated
with the luminal circumference, indicating that the sensation during antral distension depends
on deformation rather than on tension.
Biomechanical Properties of Normal Small Intestine
For a century it has been commonly recognized that two fundamental patterns of motility
are conducted by the intestine [Szurszewsky, 1998]. 1. Propulsion- peristalsis: A ring of
muscle contraction appears on the oral side of a bolus of ingesta and moves toward the anus,
propelling the contents of the lumen in that direction. As the ring moves, the muscle on the
other side of the distended area relaxes, facilitating smooth passage of the bolus. 2. Mixing-
Segmentation contraction: This contraction is a common type of mixing motility seen
especially in the small intestine. Segmental rings of contraction chop and mix the ingesta.
Like other parts of the GI tract the intestine is subjected to changes in wall stress and strains
caused by these two kinds of contraction.
Using IP with and without combing ultrasound, it was demonstrated that the intestinal
wall tension-strain or stress-strain distributions showed an exponential behavior [Jorgensen et
al., 1991; Gregersen et al, 1992a; Storkholm et al, 1995; Jorgensen et al, 1995; Duch et al,
1996; Gao et al, 2003; Frokjaer et al, 2006b]. The stiffness was different between the
duodenal, jejunal and ileal segments [Storkholm et al., 1995; Duch et al., 1996]. It is now
well known that the biomechanical analysis should be referenced to the zero-stress state
[Gregersen et al., 2000]. Gregersen et al [1999] first time studied morphometry and strain
distribution in guinea pig duodenum with reference to the zero-stress state. Afterwards,
several studies were carried out to analyze the biomechanical properties of the normal small
intestine with reference to the zero-stress state [Gao et al., 2000a; Dou et al., 2003; Zhao et
al., 2002a]. It was found that a large axial variation of opening angle and residual strain
existed along the small intestine (Figure 4) [Zhao et al., 2002a]. The axial variation of
residual strain from duodenum to ileum correlated to the morphometric variation along the
small intestine [Gabella and Blundell, 1981].
Figure 4. Top: The distribution of the opening angle and tangent rotation angle in the rat small intestine.
Both angles increased along the length of the duodenum and had its highest value 30% down the
intestine. They decreased again further down the intestine. Bottom: Residual strains distributed at the
mucosal and serosa surface. The serosa residual strain was tensile and mucosa residual strain was
compressive with highest absolute values also close to the 30% down the intestine.
The residual strain is non-uniformly distributed across the intestinal wall, it was negative
at the mucosa surface and positive at the serosa surface [Gao et al., 2000a; Dou et al., 2003;
Zhao et al., 2002a]. Hence the serosal residual strains are tensile and the mucosal residual
strains are compressive. Distension experiments showed that the residual strain makes the
stress distribution through the wall more uniform in the pressurized state [Gao et al., 2000a].
From the duodenum data it appears that the neutral axis (where residual strain is 0) is located
closest to the mucosa. This is even more pronounced in the distal parts of the small intestine
where it is located about 20% away from the mucosal surface [Dou et al., 2003]. The
implications of these data are clear. The zero-stress state is better to be accounted for in
biomechanical studies of the small intestine. Furthermore, as it has been demonstrated in
arteries, the residual stress reduces the stress concentration at the inner wall, thereby offering
a better protection at the GI tract against injury due to contractile activity and against the flow
of luminal contents. The data of stress-strain analysis showed that the duodenal wall is stiffer
than the jejunal and the ileal wall in circumferential direction, whereas the ileum was softer
longitudinally than the proximal small intestine [Dou et al., 2003]. Furthermore, all segments
were stiffer in longitudinal direction than in circumferential direction and all curves show an
exponential behavior [Dou et al., 2003]. The exponential behavior is consistent with
mechanical studies on other biological tissues [Dobrin, 1978; Fung, 1991; Fung and Liu,
1992]. The differences in stiffness is likely associated with the specialized functions of the
proximal and distal segments of the small intestine, e.g. the duodenum has a large influence
on gastric emptying and has been proposed to act as a capacitative resistor whereas the distal
ileum acts as a reservoir [Schulze-Delrieu, 1991]. The transit time is slower distally [Bueno et
al., 1975, 1994] which may partly be explained by the difference in the viscosity of the
chyme. However, the passive elastic properties may also contribute to differences in intestinal
flow patterns: a passing bolus will be slowed to a lesser degree in the duodenum where the
wall stiffness is high whereas the compliant ileal wall bulges leading to pooling of luminal
contents and decreased flow. The deformation of the intestine during contraction and
relaxation is three-dimensional. To obtain a more thorough understanding of the intestinal
transport function, we need to know the stresses and strains in multiple directions. In previous
studies it was observed that the intestine bends outwards in longitudinal direction when a
radial cut is made. Therefore, to obtain a more detailed description of the stress-free state of
the intestinal wall, longitudinal residual strain measurements are needed [Dou et al., 2006].
The longitudinal residual stress can be characterized by a bending angle (unit: degrees per
unit length and positive when bending outwards). The no-load state can be used as reference
for the longitudinal strain computation because the mid-wall length did not change from the
no-load state to zero-stress state in longitudinal direction. The average bending angle is in the
whole small intestine approximately 55/mm. As the circumferential residual strain
distribution, the longitudinal residual strain is tensile at the serosal side and compressive at
the mucosal side. Difference exists between the duodenum and ileum in longitudinal residual
strain at the mucosal side. It may be related to the thicker mucosal layer and faster passage of
food in the duodenum than in the ileum. By comparison, the residual strains in longitudinal
direction measured in this study are smaller than those in circumferential direction [Dou et al.,
2003], especially at the mucosal side. This could be due to that the largest tensile stress and
strain are distributed in the circumferential direction when the intestine is distended during
the passage of food. Hence, it is reasonable that larger residual strain exists in circumferential
direction. Smith et al [2003] measured the time-dependent (viscoelastic) behavior in the
change of the small intestinal opening angle and tested how well the behavior could be
described by the Kelvin model for a standard linear solid. The change of opening angle over
time for all the small intestine segments fits well to the standard linear spring-dashpot model.
This viscoelastic constant of the rat small intestine is fairly homogenous along its length. The
data obtained from this study add to a base set of biomechanical data on the small intestine
and provide a reference state for comparison to other tissues, diseased intestinal tissue or
intestinal tissue exposed to drugs or chemicals.
In the small intestine, submucosa is rich in collagen and is believed important for the
passive biomechanical properties. Fackler et al [1981] observed the effect of stress on rat and
bovine intestine using polarizing optical microscope, especially focus on the orientation of
collagen in the submucosa. Collagen fibres were found to be biaxially oriented at

approximately +30 degrees and -30 degrees to the longitudinal direction. The fibres are
arranged in layers with the fibres in each layer densely packed in parallel undulating arrays.
The initial response to stress is straightening of the fibres. Gradual straightening of the fibres
is related to the increasing stiffness of the tissue observed in the stress--strain relationship.
Once the fibres are straightened, the biaxial orientation of the fibres produces higher strength
in the longitudinal direction than in the transverse direction. Storkholm et al [1998] compared
the stress-strain distributions obtained from isolated segments of the guinea pig duodenum,
jejunum, and distal ileum, and the relation between the elastic properties and the collagen
content. An association was found between the collagen content and the incremental elastic
modulus (stiffness). Recent Chen et al [2008] first provided direct experimental evidence that
the villi are important for the biomechanical properties of guinea pig small intestine in
circumferential direction, because the villi not only affect the zero-stress state configuration
but also partially affect the stress-strain distribution in the intestinal wall (Figure 5).
Figure 5. The opening angle (top) was smaller in the segments without villi than in the segments with
villi. Stress-strain analysis (bottom) showed that the circumferential stress-strain curve of the segment
with villi located at the right compared to the segment without villi, indicating the villi do not support
tensile stretch. The curve was located in the middle if the villi were not taken into account in the
computation of the stress-strain in the intact segment.
Therefore, the villi should be taken into account in the analysis of biomechanical properties of
the intestinal wall. Li et al [2008] recently investigated the effect of smooth muscle tone on
morphometry and residual strain in rat duodenum, jejunum and ileum. It was demonstrated
that the intestinal mechanical properties are affected by intestinal SM tone. The tone can be
evaluated by measuring the opening angle and residual strains of sectors in intestinal
segments with and without SM relaxation.
Biomechanical Properties of Normal Colon and Rectum
Same as other parts of GI tract, the biomechanical properties are also important for colon
physiological function. Tensile stretch test showed that rat colon has a tensile strength of
around 50 g/mm2 and increased in strength from proximal to distal [Watters et al.,1985a].
Later, Watters et al [1985b] compared the features of the African and European colon in vitro.
The tensile strength in a Kampala group was greater than in an Edinburgh one. The width at
burst was greater in the Kampala group and declined with age. Stress-relaxation was similar
in both groups. In view of the similar properties in childhood of colons from Edinburgh and
Kampala, the strength of the adult African compared with European colons may derive later
from environmental factors such as diet.
The quasi-linear viscoelastic law could be applied to the in vivo determination of the
mechanical properties of the human rectal wall [Bouchoucha et al., 1994]. It was
demonstrated that the shapes of the pressure-relaxation curves at different level of distension
were significantly different. Tonic and phasic colonic motility and compliance of the
transverse and sigmoid colon were assessed using a combined barostat-manometry assembly
in healthy subjects [Ford et al., 1995]. They demonstrated that there are quantitatively
different but qualitatively similar phasic and tonic responses to the meal in the two colonic
regions. Differences in the viscoelastic and luminal dimensions may partly account for these
differences in tonic responses. Bharucha et al [2001] studied the colonic viscoelastic
properties of the human descending colon by assessing pressure-volume (P-V) relationships.
Quasi-static P-V curves were approximated to a power exponential function and revealed
hysteresis, indicative of viscoelasticity.
Petersen et al [2001] compared the pain and biomechanical responses to standardized
distension of the human colon. The biomechanical properties in the sigmoid colon and rectum
were alike. For a given wall tension and circumferential strain the sensibility seems equal in
the rectum and the sigmoid colon. The observed difference in perception between the two
segments was related to the greater CSA in the rectum. Gao and Gregersen [2000b] studied
the biomechanical and morphological properties in rat large intestine with referenced to the
zero-stress state. The wall thickness and wall thickness-to-circumference ratio increased in
the distal direction. The opening angle varied along the colon with the highest values in the
beginning of the proximal colon. The residual strain at the inner surface was negative
indicating that the mucosa-submucosal layers of the large intestine in no-load state are in
compression. The stress-strain distributions were exponential. All segments were stiffer in
longitudinal direction than in the circumferential direction. The transverse colon was stiffest
both in the circumferential and longitudinal directions. Frokjaer et al [2007a] used magnetic
resonance imaging (MRI) to evaluate the three-dimensional geometry and mechanosensory
properties of the sigmoid colon. The geometry of the distended sigmoid colon was complex
and the spatial distributions of the biomechanical parameters were non-homogeneous. The
circumferential length, strain, pressure and wall stress increased as a function of bag volume.
In response to butylscopolamine, the pressure and wall stress were reduced and the stress-
strain curves were shifted to the right. The sensory response was a linear function of the
biomechanical parameters and decreased in response to butylscopolamine as a function of
volume.
Rao et al [1997] compared incremental elastic modulus (IEM) with rectal compliance
during ramp inflation. Because slope of stress-strain relation can reflect rectal wall stiffness,
the study emphasized the importance of rectal wall viscoelastic evaluation based on stress and
strain rather than compliance. IEM had a linear relation with strain in contrast to compliance.
The series elastic component (SEC), a measure of passive viscoelasticity, was rate
independent elasticity and a rate dependent viscosity. Qiao et al [2005] showed that the
mechanical properties of rectal wall exhibited pronounced nonlinear and anisotropic
characteristics. However, Dall et al [1993] demonstrated that the circumferential wall tension
of human rectum increased in a linear way. Rectal compliance decreased in a non-linear way
with no further decline between 30 and 40 cmH2O. The pressure elastic modulus increased
steeply until a distension pressure of 35 cmH2O with no further increase to 40 cmH2O. This
suggests that rectal tone is reduced as the muscle fails to resist further distension at 35 cmH2O
and higher pressures.
Biomechanical Properties of Normal Sphincter Regions in GI Tract
Several sphincters, such as upper esophageal sphincter (UES), esophageal-gastro junction

(EGJ) including lower esophageal sphincter (LES), pyloric sphincter, sphincter of Oddi (SO),
ilealcolic sphincter, and internal and external anal sphincters (IAS and EAS) distributed along
the GI tract. UES and LES control entry into the esophagus and passage into the stomach, the
latter normally prevents reflux of gastric contents. The pyloric sphincter controls passage into
the duodenum and the ileocolic sphincter controls movement between the ileum and caecum.
The SO regulates bile and pancreatic flow into the duodenum, makes diversion of hepatic bile
into the gallbladder and prevents reflux of duodenal contents into the pancreaticobiliary tract.
The IAS and EAS control stool out of the body. Therefore, the mechanical properties to
maintain the opening and closing of the sphincters are important for their physiological
functions.
Studies with biomechanically analyzed the mechanism of UES opening showed that the
swallow-related UES opening are subject to volitional augmentation [Kahrilas et al., 1991].
Jacob et al [1989] showed that trans-UES transport of increasing swallow bolus volumes is
accomplished by modulating sphincter diameter, opening interval, and flow rate. High-
resolution manometry could accurately, comprehensively, and highly reproducibly depict the
pharyngo-esophageal (PE) segment space-time-pressure structure and specific physiological
events related to UES opening and transsphincteric flow during normal swallowing [Williams
et al., 2001]. Intrabolus pressure variations are highly dependent on position within the PE
segment and time. Biomechanical and bolus flow aspects of swallowing changed when
people performed effortful swallows with boluses [Hind et al., 2001]. Using simultaneous
high-resolution endoluminal sonography and manometry, Miller et al [2004] evaluate the
motion, morphology, and pressure of the UES. The UES has a greater muscular CSA and
resting pressure than the upper esophageal body. In the cadaver studies, the UES was imaged
in conjunction with a significant increase in pressure, indicating that the pressure is due to
passive mechanical conformational changes.
Force-velocity analysis [Cohen and Green, 1973]demonstrated that both the work and
power generated during esophageal and LES muscle contraction were determined by: (a) the
initial muscle length produced by the preload, (b) the afterload against which the muscle was
contracting, and (c) the contractility of inotropism of the muscle. Biancani et al [1973]
showed that pressure-diameter curve of LES was sigmoid in shape; at first it showed a decline
and then an increase followed by decline in pressure again as function of increasing inside
diameter. The ratio of wall thickness to inside radius varied with inside diameters and this
ratio increased steeply at small inside diameters. The tension diameter curves of the sphincter
muscle showed that optimal tension development occurred not near sphincter closure but at a
much wider diameter and that this muscle developed tension even at small luminal diameters.
Massey et al [2006] demonstrated that transient LES relaxations must occur before the
gastroesophageal junction can open. Swallow and esophageal distension can induce the
relaxations of the LES due to axial stretch [Dogan et al, 2007]. Accurate monitoring of LES
and GEJ pressures is important for the diagnostic assessment of many patients and for
advancing research into several aspects of the mechanical function of the LES and its
surrounding structures [Dent, 2007]. Recently, McMahon et al [2007a] evaluated the function
of the EGJ using a functional lumen imaging probe (FLIP). The FLIP measured eight cross-
sectional areas (CSAs) 4 mm apart together with the pressure inside a saline-filled cylindrical
bag. The geometric profile of the EGJ was reconstructed in a video animation of its dynamic
activity. A plot of curve-fitted data for the smallest CSA vs. pressure after balloon distension
indicated that the pressure increased from 18 cmH2O at a CSA of 38 mm2 to a pressure of 37
cmH2O at a CSA of 230 mm2 for the healthy controls. Therefore, FLIP represents the first
dynamic technique to profile the function and anatomy of the EGJ.
In vitro force-length curves of rings from the gastroduodenal junction were showed that
basal, total, passive, and active forces were higher for pyloric rings than for duodenal and
antral rings [Biancani et al.,1980]. Furthermore, in the pylorus the force-length curves were
shifted to the left along the length axis due to a narrowing at the gastroduodenal junction.
These data indicated that the higher forces and pressures observed in the pylorus are due to a
combination of higher active and passive components. Keinke et al [1984] showed in a dog
study that gastric emptying is controlled by (1) the depth of the antral waves, (2) the pyloric
sphincter opening, (3) the receptive relaxation of the duodenum and (4) the type of the
duodenal contractions. Shirazi et al [1984] showed in a cat study that duodenum provided a
resistance to gastric emptying over and above that generated by the pylorus, and that the
duodenal resistance is caused by limitations in duodenal capacity. The mechanical activity of
the isolated stomach and duodenum is affected by the chemical composition of their luminal
contents. Mandrek and Kreis [1992] showed that mechanical responses of gastric and pyloric
smooth muscles to acetylcholine, histamine, substance P, noradrenaline and adrenaline were
different. The spatial patterning of antropyloric pressure waves sequences is diverse [Sun et
al., 1997], and may explain the differing mechanical outcomes among individual gastric
contractions. Shafik [1998] showed that pyloric sphincter contraction and antral dilatation
depended on duodenal distension, suggested duodenopyloric reflex is existed. This reflex
appears to prevent duodenopyloric reflux. Indireshkumar et al [2000] demonstrated that
gastric emptying of nutrient liquids is primarily through the "pressure pump" mechanism
controlled by pyloric opening during periods of relative quiescence in antral contractile wave
activity.
The Intramural neural control and motility of the SO in opossum were studied by Helm et
al [1985, 1989]. The SO is innervated by intramural cholinergic excitatory nerves and
nonadrenergic noncholinergic inhibitory nerves. The frequency of SO phasic contractions is
determined by control wave frequency. Mechanical properties of isolated smooth muscle
from rabbit SO and duodenum were studied by Elbrønd et al [1991]. The difference was
found between circular smooth muscle strip of duodenum and the SO. Measurements of SO
electrical and mechanical activity are feasible and can be done as part of ERCP [Abell et al,
1998]. The frequency (cycles/min) of electrical signal was about 4.9 and tonic pressure was
about 15 mmHg in SO. It is important to study the dynamic changes of geometry and pressure
in the SO associated with other parts if GI tract, such as the stomach, duodenum and bile
ducts.
The associations existed between volume of rectal distension and the parameters of IAS
relaxation [Penninckx et al., 1992]. Because the IAS ring cannot be completely closed, the
anal mucosa and the haemorrhoidal plexuses fill the gap. By compressing these tissues, the
IAS perfectly closes the anal canal to retain not only solids but also fluid stool and gas. Acute
rectal distension and rectal activity, mainly through intramural pathways, induce reflex IAS
relaxation, permitting the rectal contents to be sampled by receptors in the upper anal canal
while continence is temporarily maintained by EAS activity and by expansion of the
haemorrhoidal cushions. Glavind et al [1993] studied the mechanical properties of isolated
smooth muscle from human rectum and IAS. The muscle resting active tension was different
between rectum and IAS. The mechanical responses of smooth muscle in the IAS were
modulated by alpha-adrenergic excitatory and non-adrenergic, non-cholinergic inhibitory
nerves. During defecation, the relaxation of IAS smooth muscle was associated with a
characteristic motility pattern of the colon and anorectum [Matsufuji and Yokoyama, 2003].
A three-dimensional (3-D) physiological model of the IAS bioengineered in vitro from
isolated smooth muscle cells was developed by Hecker et al [2005]. The bioengineered
sphincter rings showed striking functional differences when compared with bioengineered
rings made from isolated colonic smooth muscle cells. The EAS is a triple-loop system; each
loop can function as a separate sphincter through voluntary inhibition action and mechanical
compression [Shafik, 1987]. Tension-stress can effect the growth of the EAS [Li et al., 1995]
Summary of the Biomechanical Properties of GI Tract
As mentioned above the GI wall has unique and complex mechanical properties
[Gregersen and Kassab, 1996]: (1) anisotropic properties. The GI wall has complex three-
dimensional structures with different material properties in different directions. This feature is
called anisotropy and it implies that a large set of material constants have to be specified in
order to completely describe the mechanical behaviour. (2) non-linear stress-strain
relationship. The non-linear mechanical behaviour between stress and strain facilitates stretch
in the physiological pressure range and prevents overstretch and damage to the tissue at
higher stress levels. Owing to the non-linearity of the material properties it is necessary to use
more complex constitutive equations than Hooke's law. (3) large strain (finite deformation).
This means that large deformation elastic theory based on non-linear principles should be
applied to the study of GI mechanics. (4) high degree of viscoelasticity, including the time-
dependent nature of the stress-strain properties of materials. Furthermore, knowing the zero-
stress configuration is essential in any mechanical analysis since it serves as the reference
state for computing stress and strain under physiological and pathophysiological conditions.
BIOMECHANICAL REMODELLING OCCUR IN THE GI TRACT

Data on the biomechanical properties are crucial for the understanding of the dysfunction
due to disease because, 1) peristaltic motion that propels the food through the GI tract is a
result of interaction of the passive and active tissue forces and the hydrodynamic forces in the
food bolus and 2) remodelling of the mechanical properties reflects the changes in the tissue
structure that determine a specific sensory-motor dysfunction.
Figure 6. Disease induced GI remodeling in animal models.
Human studies documented that diabetes mellitus (DM) [Frokjaer et al., 2007b] and
systemic sclerosis [Pedersen et al., 2003] induced biomechanical GI remodeling. Using
different animal models (Figure 6) we have demonstrated that biomechanical and
histomophological remodeling occurred in the GI tract duo to normal physiological growth
[Gregersen et al., 2004; Lu et al., 2005], malnutrition [Dou et al., 2001, 2002a], inflammation
[Yang et al., 2009], obstruction [Gregersen et al., 1992b., Stockholm et al., 2007, 2008],
bowel resection [Dou et al., 2002b], DM [Jørgensen et al., 2001; Liao et al., 2006a; Yang et
al., 2004b, 2006; Zhao et al., 2002b, 2003a, 2003b, 2006, 2007b, 2008b], radiation injury
[Gregersen et al., 2002c], collagen changes [Gregersen et al., 2001; Fan et al., 2005] and EGF
treatment [Liao et al., 2003b; Yang et al., 2003a, 2003b; Zhao et al., 2002c, 2003c].
As mention above, the morphometric properties are best described at the zero-stress state
where no internal or external forces deform the tissue. With referenced to the zero-stress state,
combined the morphometry data and pressure data, we can computer the stress-strain
relationship of the GI wall. The stress-strain distribution mainly reflects the elastic properties
of the GI tract. Therefore, we consider the opening angle of the zero-stress state, residual
strain and stress-strain relationship as the most relevant biomechanical parameters to describe
diseases causing GI remodelling. Generally diseases and factors inducing tissue overgrowth,
such as DM, obstruction and EGF treatment, increase GI wall stiffness whereas factors
reducing tissue growth, such as fasting and low protein diet decrease GI wall stiffness (Table
1).
Table 1. Disease caused biomechanical remodeling of GI tract
Diseases Species Test Biomechanical

organs remodeling
OA RES Stiffness
Human Esophagus ND ND Circ NC Long
Duodenum ND ND Circ NC Long
Type I DM
Rat Esophagus   Circ Long ND
Duodenum   Circ Long
Jejunum   Circ Long
Ileum   Circ Long
Colon Circ Long
Type II DM Rat Esophagus   Circ Long ND
Stomach ND ND Circ
Systemic sclerosis Human Duodenum ND ND Circ
Ulcerative colitis Mice Colon   Circ Long
Fasting Rat Duodenum   Circ Long
Jejunum   Circ Long
Ileum   Circ Long
Low protein diet Mink Duodenum NC NC Circ Long NC
Jejunum   Circ Long NC
Ileum  Circ Long NC
Partial Obstruction Opossum esophagus ND ND Circ Long ND
Guinea pig Jejunum   Circ Long ND
Osteogenesis imperfecta Mice Esophagus   ND
Irradiation Mice Rectum   ND
Small bowel resection Rat Jejunum   NC
Ileum   NC
EGF treatment Rat Esophagus   Circ Long ND
Duodenum   Circ Long NC
Jejunum   Circ Long
Ileum   Circ Long
Colon   ND
Notes: OA: opening angle, RES: residual strain, Circ: circumferential direction, Long: longitudinal
direction, ND: not done, NC: no change.
The more collagen in the GI wall, the stiffer the wall is [Pedersen et al., 2003], and visa versa.
However, the effect of different factors on the opening angle and residual strain of GI tract
depend on the changes of the layered structure. Fung's hypothesis of non-uniform remodelling
states that if the inner wall grows more than the outer wall, the opening angle will increase
[Fung, 1993] whereas if the outer wall grows more than the inner wall, the opening angle will
decrease. The biomechanical remodelling in the GI tract reflects the histological structural
remodelling.
Esophageal Biomechanical Remodeling
During normal growth, the esophageal changed its biomechanical properties [Gregersen
et al., 2004]. The weight per unit length, the wall cross-sectional area, the inner and outer
circumferential length of the mucosa and muscle, and the thickness and area of the layers
increased as function of age. The opening angle was approximately 140 degrees at age 1 and
2 weeks and gradually decreased to approximately 80 degrees after 16 weeks. The stress-
strain analysis showed that the esophagus become more compliant during the first 4 weeks of
life in circumferential direction and the first 16 weeks of life in longitudinal direction (Figure
7). Rao et al [2003] showed in human study that aging was associated with a larger lumen and
a stiffer but less sensitive esophageal wall. EGF Treatment induced pronounced growth of rat
esophagus morphometrically and histologically [Zhao et al., 2003c]. At the same time, the
biomechanical properties were remodeled. The opening angle and residual strains increased
with the highest value after 14 days EGF treatment. The change in opening angle depended
largely on the change in mucosa thickness. Furthermore, the circumferential stiffness of the
esophagus reached a maximum after 7 days EGF treatment [Zhao et al., 2003c].
Biomechanical remodeling of the rat esophagus occurs in experimental type-1 DM [Yang
et al., 2004b]. Following the development of DM, for 1 week the circumferential stiffness
increased and after 4 weeks the longitudinal and shear stiffness increased. Furthermore, the
biomechanical properties of the layered esophagus were also remodeled in experimental type-
1 DM [Yang et al., 2006].
DM induced pronounced increase in the outer perimeter, inner perimeter and lumen area
in both the muscle and mucosa-submucosa layer. The growth of the mucosa-submucosa layer
was more pronounced than the muscle layer. Furthermore, the circumferential stiffness of the
mucosa-submucosa layer increased 28 days after STZ treatment. Zhao et al [2007b] recently
demonstrated that the histomorphometric and biomechanical remodeling is a feature of the
esophageal wall in type 2 diabetic GK rats. DM induced pronounced increasing of the wall
thickness and wall cross-sectional area. Histologically, the circumference, thickness, and area
of the muscle layer and the collagen fraction of mucosa-submucosa layer were significantly
increased in the GK rats. The opening angle and outer residual strain decreased significantly
in the GK rats. Furthermore, the circumferential stiffness of the esophagus is significantly
higher in the GK rats than in the normal rats. Frokjaer et al [2007b] showed in human study
that the esophageal wall and especially the mucosa-submucosa layer had increased thickness
in the diabetic patients and the longitudinal and radial compressive stretch was less in
diabetics.
6
1 week
2 weeks
5
Circumferential Stress (kPa)

4 weeks
8 weeks
4
16 weeks
24 weeks
3 32 weeks
0
-0.5 0 0.5 1 1.5 2
Circumferential Strain
6
5
Longitudinal Stress (kPa)
0
0 0.5 1 1.5 2
Longitudinal Strain
Figure 7. The graphs show the stress-strain relations for circumferential (top) and longitudinal direction
(bottom). The circumferential and longitudinal stress-strain curves were exponential. Circumferential
stress-strain curves shifted from left to the right up to four weeks of age where after no further change
was observed. Longitudinal stress-strain curves shifted from left to the right up to 16 weeks of age
where after no further change was observed. The esophageal wall was stiffer in longitudinal direction
than in circumferential direction.
The rabbit esophageal wall with varices showed a significant loss of distensibility and
recovery of the wall distensibility when the varices were treated [Gregersen et al., 1988]. The
partially obstructed opossum esophagus expressed dilatation and decreased ability of the
esophageal wall to resist intraluminal pressure [Gregersen et al., 1992b]. By means of IP, Juhl
et al [1994] showed that endoscopic sclerotherapy entailed an inversion of the normal
esophageal configuration, with narrowing 5 cm above the GEJ and increased CSAs 10 cm
above the GEJ. Gregersen et al [2001] investigated esophageal morphometry and residual
strain in a mouse model of osteogenesis imperfecta murine (oim) mice with collagen
deficiency. Compared with wild-type mice, the esophagi in oim mice had smaller layer
thicknesses and areas with the largest reduction in the submucosa, larger opening angles, and
larger residual strains at the serosal and mucosal surfaces and the mucosa-submucosal-muscle
layer interface. The data suggested that the residual stress in esophagus is caused by the
tension in the muscle layer rather than the stiffness of the submucosa in compression and that
the remodeling process in the oim esophagus is due mainly to morphometric and
biomechanical alterations in the submucosa.
Holland et al [1993] studied esophageal compliance in naturally occurring canine
megaoesophagus. In most cases of canine megaoesophagus the viscoelastic properties of the
esophageal wall are significantly altered, however in such cases the disorder is unlikely to be
purely dynamic and that processes other than the duration of esophageal dilatation are
responsible for the alteration in esophageal wall biomechanical properties. Rao et al [1996]
showed that the esophageal compliance was reduced in unexplained chest pain patients.
Sensory and biomechanical properties of the esophagus in non-erosive reflux disease (NERD)
were studied by Reddy et al [2007]. The NERD patients were less sensitive to mechanical
stimulation as assessed by the cross-sectional area and volume and hypersensitive to heat
stimuli of the esophagus, with an increase in referred pain to the evoked visceral pain. The
data indicate that peripheral sensitization of heat-sensitive pathways together with facilitation
of central pain mechanisms is important in the pathogenesis of NERD. Remes-Troche et al
[2007] demonstrated that the esophagus was more reactive and less distensible in
gastroesophageal reflux disease patients than controls. After Nissen fundoplication, in
patients with persistent symptoms, the sensory thresholds were unchanged but the esophageal
wall reactivity decreased and distensibility improved.
Stomach Biomechanical Remodeling
Gastroparesis is one of the most common motility disturbances in DM and it affects up to

58% of diabetics [Horowitz et al., 1991; Malagelada et al., 1980; Verne et al., 1998].
However, the etiology is poorly understood. Patients with diabetic gastroparesis have delayed
gastric emptying and abdominal pain [Rothstein, 1990], suggesting that the sensitivity of
gastric mechanoreceptors, gastric accommodation, and the mechanical properties may be
important in the etiology of symptoms [Moragas et al., 1993; Rayner et al., 2000; Schwizer et
al., 2002]. Barostat studies have shown that the compliance in the proximal stomach may be
altered by DM [Karakida and Homma, 1989; Noriyuki et al., 1999; Rayner et al., 2000]. More
recently, Kumar et al [2008] performed intermittent, phasic balloon distensions of the
stomach using a gastric barostat device in diabetic patients and normal subjects. The
accommodation response to a meal was significantly impaired in patients with DM when
compared to controls. However, several studies have demonstrated that perception of gastric
distension depends on wall tension rather than on intraluminal pressure or volume [Gregersen
et al., 1992c, 1992d; Distrutti et al., 2000]. Hence evaluation of tension and its surface
variation seem to be the best approach in designing gastric sensitivity and function studies.
Liao et al [2006a] made three-dimensional geometry analysis of the stomach in type II
diabetic GK rats. The diabetic stomach exhibited lower material compliance not only in the
proximal stomach but also in the entire stomach. The biomechanical properties differed
between regions and in different directions in both the non-diabetic and the diabetic stomach.
Therefore, the changes due to DM may contribute to the delayed gastric emptying and the
symptoms in diabetic patients.
Salet et al [2006] studied the responses to gastric distension in functional dyspepsia.
Patients with functional dyspepsia show slightly higher compliance to mechanical distension.
Their visceral perception of mechanical stress is enhanced. In contrast with the balloon
distension, relaxation after a meal was less.
Intestinal Biomechanical Remodelling
Intestinal physiological growth induced morphometric and biomechanical remodeling

[Lu et al., 2005]. Small intestinal mass and dimensions increased many-fold with age. The
opening angle of duodenum was approximately 220 degrees and 290 degrees during the first
and second week after birth and decreased to 170 degrees at other ages. The opening angle of
ileum ranged between 120 degrees and 150 degrees. The residual strain of duodenum at the
mucosal surface did not vary with age whereas the residual strain of ileum at the mucosal
surface decreased with age. The stress-strain analysis showed that the small intestine became
more compliant during early life.
Vinter-Jensen et al [1996] demonstrated that EGF treatment for 1 to 4 weeks caused a
time-dependent increase in intestinal weight. The growth was characterized by increased wall
thickness, increased cross-sectional area and reduced wall stiffness. Zhao et al [2002c]
demonstrated that EGF treatment was associated with pronounced morphometric and
histological changes in the rat small intestine. The opening angle and residual strain increased
in all segments of the small intestine during EGF treatment with the highest value in the 14
days EGF treatment group. Linear regression analysis demonstrated that the opening angle
mainly depended on the mucosa thickness and area. Furthermore, the circumferential stiffness
increased in the duodenum and decreased in the jejunum and ileum during EGF treatment. A
plateau was reached after 7 days where after it started to normalize. In the longitudinal
direction, all intestinal segments became stiffer after EGF treatment for 7 days. After 14 days
the curve started to normalize in duodenum and jejunum but not in the ileum. Yang et al
[2003a] also showed that EGF caused significant changes both in the morphology and in the
passive mechanical properties of the rat ileum. Liao et al [2003b] has shown that the
longitudinal modulus of rat intestine after 4 days EGF treatment was significantly larger than
in a control group. The cross modulus decreased during the first 4 days of EGF treatment
thereafter it increased to a maximum at 7 days.
Nutrition plays an important role in the maintenance of normal gut structure and function.
The small intestine undergoes functional changes when food is withheld. To explore the
morphological and biomechanical remodeling during starvation, intestinal segments from the
fed and fasted rat duodenum, jejunum, and ileum were investigated by Dou et al [2002a].
Fasting decreased the outer circumferential length, wall thickness, wall area, inner
circumferential length, and luminal area in the three intestinal segments. Histological
examination showed that the mucosal and the submucosal thickness decreased during fasting,
whereas the muscle layers were unchanged. Fasting shifted the stress-strain curves to the right
in both circumferential and longitudinal directions at the three segments, indicating that the
intestinal wall became soft due to fasting. However, the mechanism behind the fasting
remodeling is not yet clear. In order to increase our understanding about the mechanism of the
effect of fasting on GI remodeling, Chen et al [2008] studied the intestinal remodeling in

mink fed with reduced protein content. Feeding the low protein diet induced decreasing the
wet weight per unit length, wall thickness and area. Histologically, the wall thickness, mainly
the mucosa, was decreased in the jejunum and ileum. The stress-strain curves were shifted to
the right in the circumferential direction in all three intestinal segments, indicating decreased
stiffness of the wall in the circumferential direction. This study demonstrated that feeding low
protein diet in mink induces histomorphometric and biomechanical remodelling in intestine
comparable to the previously observed changes during fasting (Figure 8). Therefore, the study
provided evidence that protein malnutrition is an important reason for fasting-induced
intestinal remodelling. In order to study the recovery of starvation-induced structural and
biomechanical remodeling in the rat small intestine, Dou et al [2001] studied the effect of re-
feeding on the biomechanics of fasting intestine. Fasting-induced biomechanical and
structural remodeling could be normalized by re-feeding in a time- and location-dependent
way.
Figure 8. The relationship between circumferential and longitudinal stress and strain. In the
circumferential direction, the stress–strain curves in all segments of the L group were shifted to the
right compared with those of H and M groups. Thus, the stiffness of intestinal wall decreased (all
P<0.01 compared with normal group). No significant difference was observed in the longitudinal
direction in all three intestinal segments.
Many human [Folwaczny et al., 1999; Charlton et al., 2000; He et al., 2001; Secondulfo
et al., 2004] and animal [Folwaczny et al., 1999; Verne and Sninsky, 1998; Zoubi et al., 1995;
Bhor et al., 2004; Noda et al., 2001; Tormo et al., 2002; Adachi et al., 2003] studies have
shown that DM causes morphological alterations in the small intestine. However, only few
data exist relating to biomechanical remodeling of the small intestine in DM. Jorgensen et al
[2001] demonstrated that the intestinal wall stiffness increased during the development of
STZ-induced rat DM. Zhao et al [2003a] studied the biomechanical and morphometric
intestinal remodeling during experimental DM in rats. The blood glucose concentration
increased four- to fivefold in the diabetic rats. STZ-induced DM generated pronounced
increase in the weight per centimeter length, wall thickness and wall cross-sectional area in
all intestinal segments during DM. Histological analysis showed that the thickness of the
intestinal layers was increased in all segments during DM. In the duodenum the opening
angle and residual strain did not change in the first 2 weeks and decreased after 4 weeks. In
the jejunum and ileum the opening angle increased after 1 week in the diabetic group.
Furthermore, it was found that the circumferential and longitudinal stiffness of the intestinal
wall increased with the duration of DM. Moreover, Zhao et al [2003b] showed the
viscoelastic behavior of rat intestinal wall was also changed, i.e., stress relaxation decreased
during the development of DM. Gliclazide [Zhao et al., 2002b] and KYQWJJ [Sha et al.,
2006] treatment could partly restore the changes of blood glucose level and the remodeling of
morphometry and residual strain of small intestine in diabetic rats. The linear regression
analysis demonstrated that the effect of Gliclazide and KYQWJJ on intestinal opening angle
and residual strain is partially through its effect on the blood glucose level.
Small intestinal obstruction is a frequently encountered clinical problem. To understand
the mechanisms behind obstruction and the clinical consequences, data are needed on the
relation between the morphologic and biomechanical remodeling that takes place in the
intestinal wall during chronic obstruction. Stockholm et al [2007] demonstrated that the
luminal cross-sectional area, wall thickness, and elastic modulus in circumferential direction
increased in a time-dependent manner proximal to the obstruction site. The circumferential
stress-strain curves of the proximal segments in 2- and 4-week groups shifted to the left,
indicating the intestinal wall became stiffer. Histological examination revealed a massive
increase in the thickness of the muscle layer especially the circular smooth muscle layer. The
collagen content proximal to the obstruction site was significantly larger in the partially
obstructed animals compared to controls. Strong correlation was found between the collagen
content and the elastic modulus at stress levels of 70 kPa and 10 kPa proximal to the
obstruction site suggesting that alteration of collagen has great impact on the mechanical
remodeling.
Dou et al [2002b] studied morphometric and biomechanical remodeling in the intestine
after small bowel resection in the rat. It was demonstrated that resection of the majority of the
small bowel results in significant remodeling in structural and residual strain properties in the
rat small intestine. The remodeling seems to be guided by the need for a greater absorptive
surface area rather than for a change in the stress-strain properties. Abnormalities of the small
intestine have also been indicated in systemic sclerosis. Pedersen et al [2003] showed in a
human study that systemic sclerosis resulted in increased stiffness and impaired muscle
function of the duodenum.
Colon and Rectal Biomechanical Remodelling
The elasticity of the rectal wall after accommodation to distension was significantly
greater in patients with Hirschsprung's disease [Arhan et al., 1978], and the time taken by the
rectum to accommodate was longer. The increased elasticity correlated well with severity of
the illness. The quasilinear viscoelastic law was applied to the in vivo determination of the
mechanical properties of the rectal wall in controls and in children with constipation by
Bouchoucha et al [1997]. The elastic response was similar in the 2 groups. The reduced
relaxation function was significantly different between the 2 groups, with the absence of an
asymptotic value in the group with constipation. The effects of a diverting colostomy on
weight, collagen content, and biomechanical strength of the left colon was investigated in rats
by Kissmeyer-Nielsen et al [2003]. It demonstrated decrease in collagenous proteins and
mechanical strength of distal colon after diverting colostomy in rats.
Radiotherapy of malignancies in the pelvis is associated with both early and late gut
reactions with subsequent alterations in rectal function because this part of the gut often is
included in the radiation field. Therefore, the biomechanical properties of rectum may be
affected by irradiation. Early and late effects of irradiation on morphometry and residual
strain of mouse rectum were studied by Gregersen et al [2002c]. The irradiated groups had
25-75% smaller opening angles than the normal groups. The early irradiated group had higher
values of residual strain than the control group. Furthermore, proximal to the irradiated zone
in the late irradiated group had significantly higher values of residual strain compared to the
normal group, indicating a higher stress level proximal to the irradiated zone. Therefore,
radiation therapy remodelled the morphometry and zero-stress state of rectum.
Age-related alterations in the strength and collagen content of left colon in rats were
studied by Christensen et al [1992a]. An accumulation of collagenous proteins was found in
old rats compared with middle-aged rats and this was accompanied by a decrease in the
strength, which may deteriorate the functional integrity of the left colonic wall with age. The
effect of growth hormone treatment on the left colon was investigated in 4-month-old Wistar
rats by Christensen et al [1990]. Growth hormone increased the mass, the collagenous
proteins, and the strength of rat colon. Studied in 27-month-old male rats showed that the
maximum load and stiffness of the left colonic specimens in the b-hGH-treated group were
enhanced by 27 and 60%, respectively, and the strain at maximum load was 24% less than in
the control group [Christensen et al., 1992b]. The systemic EGF treatment caused
remodelling of the morphology of the zero-stress state in the large intestine in a time-
dependent manner [Yang et al., 2003b]. The growth was most pronounced in the ascending
and transverse colon and involved mainly the mucosal layer.
The pain, urgency, and incontinence in ulcerative colitis may be related to changes in
viscoelastic properties of the gut wall or to alterations of the sensory pathways. The profound
organic changes in UC indicate that the mechanical and mechanosensory functions are
affected. Reduced compliance and increased sensation have been reported in several clinical
studies [Rao et al., 1987; Loening-Baucke et al., 1989]. Used an advanced rectal probe,
Drewes et al [2006] studied the mechanosensory and smooth muscle properties in patients
with active ulcerative colitis. There were no differences in compliance or stiffness between
the patients with active ulcerative colitis and normal subjects, but the hypersensitivity was
abolished after muscle relaxation. This finding demonstrated that the smooth muscles were
tonically contracted in the inflamed rectum, resulting in a decreased rectal circumference.
Recently Yang et al [2009] demonstrated that Oxazolone-induced colitis in BALB/C mice

induced increasing wall thickness, opening angles, circumferential and longitudinal wall
stiffness in the colon. Furthermore, the mechanical remodelling was associated with the
morphometric and histological changes (Figure 9).
Figure 9. Stress-strain data for both circumferential and longitudinal direction for normal and UC mice
(top). The UC mice were stiffer for both directions (P<0.01). Similarly, the circumferential and
longitudinal constants increased in the UC colon (middle). Association between a11 and the degree of
inflammation were found (bottom).
Stiffness of the colon wall will lead to increased resistance to normal passage of feces
and air in the affected segment, proximal dilatation, and resulting pain and urgency [Drewes
et al., 2006].
More recently, Zhao et al [2008b] showed that biomechanical and histomorphometric
colon remodelling occurred in STZ-Induced diabetic rats. The blood glucose level increased
3-4-fold in the diabetic rats compared with the controls. DM generated pronounced increasing
weight per length, wall thickness, and wall cross-sectional area in the diabetic colon.
Histologically, the thickness of all layers increased during DM, especially the mucosa layer.
The opening angle and absolute values of residual strain increased in the diabetic group.
Furthermore, DM increased the circumferential and longitudinal stiffness of the colon wall.
The observed changes in residual strain, opening angle, and stress-strain relation may be
contributing factors to colonic dysfunction and abdominal pain in diabetic patients.
Park et al [2008] recently demonstrated that there were significant differences in static
compliance between the normal subjects and the irritable bowel syndrome (IBS) patients.
However, Drewes et al [2001] demonstrated that no differences between IBS patients and
normal subjects were seen in strain and tension of the rectum and sigmoid colon. Therefore,
they thought that visceral hypersensitivity in IBS seems to be related to alterations in the
nervous system rather than biomechanical parameters such as the tension and strain of the gut
wall.
Biomechanical Remodeling of Sphincter Regions in GI Tract
Less UES flexibility in the swallows of older men was demonstrated by Logemann et al
[2000]. The mechanical malfunction of the UES and pharynx during mastication and
swallowing were found in patients with Parkinson's disease [Johnston et al., 1995].
Manometric abnormalities of the esophageal body and LES have also been documented.
Williams et al [2002] studied the biomechanics of failed deglutitive UES relaxation in
neurogenic dysphagia. Etiology of failed UES relaxation was 56% medullary disease, 33%
Parkinson's or extrapyramidal disease; and 11% idiopathic. Adequacy of UES opening is
influenced by the degree of preservation of the pharyngeal swallow response and
hyolaryngeal traction. In contrast, the stenotic UES displays a static loss of compliance, only
apparent once the limit of sphincter expansion is reached.
Biancani P et al [1984] showed that in vitro mechanical properties of the LES are
affected by induction of esophagitis. Sixty percent of the patients with documented increased
esophageal exposure to acid gastric juice had a mechanically incompetent LES [Zaninotto et
al., 1988], indicating the important of LES mechanical disorder in gastro-esophageal reflux
disease (GERD). The EGJ of GERD patients with hiatus hernia was more distensible and
shorter than normal subjects [Pandolfino et al, 2002]. These findings partially explain why
hiatus hernia patients are predisposed to reflux by mechanisms other than transient LES
relaxations, sustain greater volumes of refluxate, and have a reduced ability to discriminate
gas from liquid reflux. Achalasia is considered a primary motility disorder confined to the
esophagus. The LES in achalasia is frequently hypertonic and manifests absent or incomplete
relaxation in response to deglution. On the other hand, the LES and the proximal stomach act
physiologically as a functional unit whereby relaxation of the LES during deglution is
associated with receptive relaxation of the proximal stomach. Mearin et al [1995] showed that
the proximal stomach relaxation was impaired in patients with achalasia. Furthermore,
Mearin et al [2000] investigated the mechanical properties of the GEJ comprising
intraluminal pressure (measured by manometry), resistance to flow and compliance
(measured by resistometry) in health, achalasia, and scleroderma. They found that GEJ
resistance to flow is significantly increased in achalasia and decreased in scleroderma while
GEJ compliance is diminished in achalasia and normal in scleroderma. Therefore, achalasia is
a disease characterized by increased GEJ resistance and rigidity. McMahon et al [2007a] used
FLIP to measure the cross-sectional areas in EGJ and showed that in one achalasia patient
(unsuccessfully treated with dilations), the CSA never rose above the minimal measurable
value despite the pressure increasing to 50 cmH2O. In another achalasia patient (successfully
treated with dilations), the pressure only reached 15 cmH2O despite opening to a CSA of 250
mm2. Therefore, The FLIP can be used practically to evaluate difficult cases of EGJ
dysfunction and may provide a role in evaluating patients before and after therapies for
diseases affecting the EGJ such as achalasia and GERD. Severe erosive reflux disease and
Barrett's esophagus are associated with significantly greater loss of the mechanical antireflux
barrier as reflected in the presence of hiatal hernia and LES measurements [Lord et al., 2008].
Diabetes mellitus has been associated with a variety of gastrointestinal motor
disturbances. Mearin et al [1986] showed that pyloric dysmotility was existed in some
patients with diabetes. The cause of sphincter distortion in rectal prolapse is unknown but
may be a response to increased mechanical stress placed on the sphincter from the prolapse or
an abnormal response by the sphincter complex to the prolapse [Dvorkin et al, 2004].
Impact of GI Biomechanical Remodeling
It is well known that mechanosensation is importance for GI function. The

mechanosensitive nerve endings exist extensively in the GI tract where they serve a critical
role in homeostasis. The mechanosensitive afferents in the intrinsic and extrinsic pathways
were described as low-, wide-dynamic- or high-threshold tension-receptors. Therefore, the GI
tract structure as well as the stress and strain distribution in the wall is important for the GI
sensory and motor function. The GI wall structure or deformation changes in the disease will
alter the relative positions of the mechanosensitive afferents (zero setting of the
mechanosensitive afferents). The biomechanical remodelling in the disease such as alterations
of residual strain, stress distribution and wall stiffness will alter the tension and stress
distribution of the mechanosensitive afferents. Consequently, the perception and motility of
the GI tract will change as well. Therefore, the morphological changes and biomechanical
remodelling of GI tract due to diseases is likely to affect the function of mechanosensitive
afferents in the GI wall and further affect the motor and sensory function.
BIOMECHANICAL MODELING OF THE GI TRACT

Gastroenterology research has traditionally been based on experimental approaches rather
than on mathematical modelling. Most of the previous modelling efforts in the biological area
were in the cardiac and lung field but other areas are in development now. However, in the
past five to ten years several groups have independently started to model the GI tract. The use
of numerical models and, in particular, of finite element models has been extensively studied
in the field of soft tissues mechanics because of the potential they offer in the analysis of the
mechanical behaviour of morphologically complex structures, with high structural hierarchy
and constituents with non-linear behaviour [Natalia et al., 2006a, 2006b]. The effectiveness of
numerical models depends on reliable reconstructions of the morphometry of the anatomical
site under investigation and the specific loading and boundary conditions, as well as on the
definition of constitutive models capable of describing the mechanical response of the single
tissues. The large morphological complexity of the GI tract and the variability in the different
parts of the tract are well-known. The complexity increases in the characteristic folds of the
connection regions [Liao et al., 2006, 2007; Yang et al., 2007b].
The methods and current development on the biomechanical properties studies on the
normal and diseased GI tissues have been described in the above sections. Hereby, the
morphometric related modelling establishment on the GI tract will be briefly introduced.
According to the reconstruction methods of GI modelling, the establishment of the GI models
can be divided into the geometrically based models, the anatomically based models and the
theoretical analysis based models.
Geometry Based Models
The geometric modelling is the fundamental part of the GI modelling analysis. Due to the
complex geometry of the GI organs, three dimensional (3-D) models of the GI tract are most
important. The complex geometry of GI tract can be obtained by advanced imaging
techniques.
Medical image based 3-D models of in vivo GI organs have characterized the
oesophagus, stomach, small intestine, sigmoid colon, oesophageal gastric junction and rectum
based on cross-sectional imaging using ultrasonography, computed tomography (CT),
Functional Luminal Imaging Probe (FLIP) or magnetic resonance imaging (MRI) [Frokjaer et
al., 2005, 2007a; Li et al., 1994: Jeays et al., 2007; Liao et al., 2004b; McMahon et al., 2007b;
Nicosia et al., 2001]. The initial 3-D GI models were reconstructed on the basis of images
from ultrasonography, computed tomography (CT) or MRI. With the identified inner and
outer wall boundaries of each cross sectional 2-D image, a data cloud of a 3-D model was
then obtained by using a segmentation technique. With the development of the medical
devices such as IP, it is now possible to record the mechanical parameters as the luminal
pressure simultaneously as the cross-sectional medical images. Therefore, the in vivo
mechanical behaviour of the organs can be computed on the basis of the reconstructed GI
morphometric models and the recorded mechanical parameter. A reconstructed sigmoid-colon
model and the corresponding tension and stress distribution on the modelling is illustrated in
figure 10. The detailed description about the in vivo GI modelling analysis can be found in
studies of Liao et al [2004b] and Frokjaer et al [2005, 2007a].
The 3-D model reconstruction and the tension and stress indicated in figure 10 were only
based on the medical images, omitting the information about the tissue material properties
and the tissue structures. Therefore, the methods described in this section can be further
extended clinically for an individual patient based computational GI tract model analysis.
Figure 10. A sigmoid colon solid model generated directly on the basis of the 3-D data cloud (top) and
a smoothed surface model generated from the pictures re-sliced along the centre line of the solid model
(bottom). The black lines on the top panel are the slices along the centre line of the solid model.
Anatomy Based GI Modelling
For modelling analysis on the in vivo image based models, only the tension or stress was
computed on the basis of three dimensional surface geometry by using the Laplace‘s equation
and the wall thickness. The tissue structure and mechanical properties of the GI wall were
therefore not taken into account. To aid in understanding of the relationship between the
structure and function of the GI tract in health and disease states, anatomically based finite
element models are needed. The anatomically based visualization GI model is now
commercially available, however the numerical calculation based GI anatomical model
existed currently are most built by Andrew Pullan‘s group and all models were built from the
Visible Human Project. The Visible Human project provided a set of cross sectional images
of a human cadaver. On each image, data points around the interesting organ boundary were
created and then the geometry models were constructed on the basis of the distinguished data
clouds [Buist et al., 2004, 2006; Cheng et al., 2007; Lin et al., 2006a, 2006b; Pullan et a.,
2004; Yassi et al., 2004]. The anatomically based models have now been used to investigate
normal and pathological electrical activity of the stomach and small intestine [Cheng et al.,
2007; Lin et al., 2006a, 2006b; Pullan et a., 2004], the muscle functions on the gastro-
oesophageal junction during swallow [Yassi et al., 2004] and the blood flow in the mesenteric
arterial system of the human intestine [Buist et al., 2004, 2006].
Theoretical Analysis Based GI Modelling
The morphological complexity of the GI organs makes it difficult to build the

anatomically based finite element models. Hence some numerical models of the GI organs on
the basis of the theoretical analysis were built by simplifying the complex GI morphometry as
a regular geometry such as the circular cylinder for the oesophagus [Brasseur et al., 2007; Li
et al., 2006; Liao et al., 2003a; Nicosia et al., 2001; Yang et al., 2006] , sphere for the stomach
pouch [Gao et al., 2008] and some regular tubes for describing the antrodudenal junction
[Dillard et al., 2007] and esophago–gastric junction [McMahon wt al., 2007a]. Based on the
simplified model, most of the biomechanical features such as the tissue structure, tissue
properties and bolus flow thus can be expressed mathematically and the mechanical function
of the GI tract can be simulated.
Figure 11. The 3-D pouch finite element model (left) and stress (middle) and strain (right) distribution
in a pouch model before emptying.
The simplified GI tract models are now existed for describing the muscle function [Liao
et al., 2004b; Nicosia., 2001; Brasseur et al., 2007], the food transportation [Dillard et al.;
2007; Gao et al., 2008; Pal., et al., 2007; Yang et al., 2007c] and blood flow in the GI tract on
healthy and diseased situations. A simplified pouch model for describing the obesity patient
gastric emptying is illustrated in figure 11 as an example. It is indicated that the pouch
emptying curves calculated on the basis of the simplified model agrees well with the clinical
recorded results [Gao et al., 2008]
CONCLUSION
The GI tract is functionally subjected to dimensional changes. Hence, biomechanical
properties are of particularly importance. Due to the complexity of geometry, anatomy and
structure, the biomechanical properties of GI tract are anisotropic and vary along the GI tract.
Therefore, analysis of the GI biomechanical properties must take the complex geometry and
the anisotropic biomechanical properties into account. The relationship between stress and
strain is desirable for a more thorough understanding of the mechanical properties of the GI
tract. The stress-strain relationship is better to be determined in reference to the zero-stress
state. The biomechanical properties will remodel due to normal growth, surgical invention,
systemic diseases, local disorders and etc. Understanding the biomechanical properties of GI
tact and its remodeling due to different situations can help us to better understand physiology
and pathophysiology of GI tract. Based on the obtained morphometry and biomechanical
properties of the GI tract, the normal and disease remodeled GI tract can be modeled
mathematically. With the development of the imaging techniques, new analytic methods and
medical devices such as IP, the in vivo mechanical behaviour of the organs can be computed
on the basis of the reconstructed GI models. The present studies on the GI modeling can be
integrated in order to analyze complex structures for understanding biomechanical properties
and visceral perception in other visceral organs and further to be integrated as a global GI
model.
ACKNOWLEDGMENTS
NIH grant 1RO1DK072616-01A2 and Karen Elise Jensens Foundation supported this
work.
REFERENCES
Abell TL, Werkman RF, Familoni BO, Baggous W, Massie D, Vera S. Biliary, pancreatic,
and sphincter of Oddi electrical and mechanical signals recorded during ERCP. Dig. Dis.
Sci. 1998; 43: 540-546.
Adachi T, Mori C, Sakurai K, Shihara N, Tsuda K, Yasuda K. Morphological changes and
increased sucrase and isomaltase activity in small intestines of insulin-deficient and type
2 diabetic rats. Endocr J. 2003; 50: 271-279.
Arhan P, Devroede G, Danis K, Dornic C, Faverdin C, Persoz B, Pellerin D. Viscoelastic

properties of the rectal wall in Hirschsprung's disease. J. Clin. Invest. 1978; 62: 82-87.
Assentoft JE, Gregersen H, O'Brien WD Jr. Determination of biomechanical properties in
guinea pig esophagus by means of high frequency ultrasound and impedance planimetry.
Dig. Dis. Sci. 2000; 45: 1260-1266.
Azpiroz F, Malagelada JR. Physiologic variations in canine gastric tone measured by an
electronic barostat. Am. J. Physiol. 1985; 248: G229-G237.
Azpiroz F, Malagelada JR. Gastric tone measured by an electronic barostat in health and
postsurgical gastroparesis. Gastroenterology. 1987; 92: 934-943.
Barlow JD, Gregersen H, Thompson DG. Identification of the biomechanical factors
associated with the perception of distension in the human esophagus. Am. J. Physiol
Berstad A, Hausken T, Gilja OH, Hveem K, Nesje LB, Odegaard S. Ultrasonography of the
human stomach. Scand. J. Gastroenterol. Suppl 1996; 220: 75-82.
Berstad A, Hausken T, Gilja OH, Nesland A, Odegaard S. Imaging studies in dyspepsia. Eur.
J. Surg. 1998; Suppl: 42-49.
Bharucha AE, Hubmayr RD, Ferber IJ, Zinsmeister AR. Viscoelastic properties of the human
colon. Am. J. Physiol. Gastrointest Liver Physiol. 2001; 281: G459-G466.
Bhor VM, Raghuram N, Sivakami S. Oxidative damage and altered antioxidant enzyme
activities in the small intestine of streptozotocin-induced diabetic rats. Int. J. Biochem.
Cell Biol. 2004; 36: 89-97.
Biancani P, Goyal RK, Phillips A, Spiro HM. Mechanics of sphincter action. Studies on the
lower esophageal sphincter. J. Clin. Invest. 1973; 52: 2973-2978.
Biancani P, Zabinski MP, Kerstein MD, Behar J. Mechanical characteristics of the cat
pylorus. Gastroenterology. 1980; 78: 301-309.
Biancani P, Barwick K, Selling J, McCallum R. Effects of acute experimental esophagitis on
mechanical properties of the lower esophageal sphincter. Gastroenterology. 1984; 87: 8-
16.
Bouchoucha M, Jais JP, Arhan P, Landais P, Faverdin CL, Pellerin D. Variation of
rheological properties of the human rectal wall with distending volume. Clin. Invest.
Med. 1994; 17:107-114.
Bouchoucha M, Jais JP, Arhan P, Landais P, Pellerin D. Adaptation of the rectal wall to
distension in children with constipation. Clin. Invest. Med. 1997; 20: 35-40.
BrassEur. J.G, Nicosia MA, Pal A, Miller LS. Function of longitudinal vs circular muscle
fibers in esophageal peristalsis, deduced with mathematical modeling. World J.
Gastroenterol. 2007; 13: 1335-1346.
Bueno L, Fioramonti J, Ruckebusch Y. Rate of flow of digesta and electrical activity of the
small intestine in dogs and sheep. J. Physiol. 1975: 249: 69-85.
Bueno L, Fioramonti J. Neurohormonal control of intestinal transit. Reprod. Nutr. Dev. 1994;
34: 513-525.
Buist ML, Cheng LK, Yassi R, Bradshaw LA, Richards WO, Pullan AJ. An anatomical
model of the gastric system for producing bioelectric and biomagnetic fields. Physiol.
Meas. 2004; 25: 849-861.
Buist ML, Cheng LK, Sanders KM, Pullan AJ. Multiscale modelling of human gastric electric
activity: can the electrogastrogram detect functional electrical uncoupling? Exp. Physiol.
2006; 91: 383-390.
Charlton M, Ahlman B, Nair KS. The effect of insulin on human small intestinal mucosal
protein synthesis. Gastroenterology. 2000; 118: 299-306.
Chen P, Zhao J, Nielsen VH, Clausen T, Gregersen H. Intestinal remodelling in mink fed
with reduced protein content. J. Biomech. 2009; 42: 443-448.
Chen X, Zhao J, Gregersen H. The villi contribute to the mechanics in the guinea pig small
intestine. J. Biomech. 2008; 41: 806-812.
Cheng LK, Komuro R, Austin TM, Buist ML, Pullan AJ. Anatomically realistic multiscale
models of normal and abnormal gastrointestinal electrical activity. World J.
Gastroenterol. 2007; 13: 1378-1383.
Christensen H, Jørgensen PH, Oxlund H, Laurberg S. Growth hormone increases the mass,
the collagenous proteins, and the strength of rat colon. Scand. J. Gastroenterol. 1990; 25:
1137-1143.
Christensen H, Andreassen TT, Oxlund H. Age-related alterations in the strength and
collagen content of left colon in rats. Int. J. Colorectal Dis. 1992a; 7(2): 85-88.
Christensen H, Andreassen TT, Oxlund H. Increased mechanical strength of left colon in old
rats treated with growth hormone. Gerontology. 1992b; 38: 245-251.
Clapworthy GJ, Viceconti M. Seeing the EuroPhysiome: A roadmap to the Virtual
Physiological Human. ( Coordination Action #027642, STEP: A strategy for the
EuroPhysiome) . Bedfordshire: University of Bedfordshire, 2007.
Cohen S, Green F. The mechanics of esophageal muscle contraction. Evidence of an inotropic
effect of gastrin. J. Clin. Invest. 1973; 52(8): 2029-2040.
Dall FH, Jørgensen CS, Houe D, Gregersen H, Djurhuus JC. Biomechanical wall properties
of the human rectum. A study with impedance planimetry. Gut. 1993; 34: 1581-1586.
Dent J. Approaches to driving the evolving understanding of lower oesophageal sphincter
mechanical function. J. Smooth Muscle Res. 2007; 43: 1-14.
Dillard S, Krishnan S, Udaykumar HS. Mechanics of flow and mixing at antroduodenal
junction. World J. Gastroenterol. 2007; 13: 1365-1371.
Dobrin PB. Mechanical properties of arterises. Physiol. Rev. 1978: 58: 397-460.
Dogan I, Bhargava V, Liu J, Mittal RK. Axial stretch: A novel mechanism of the lower
esophageal sphincter relaxation. Am. J. Physiol Gastrointest Liver Physiol. 2007; 292:
G329-G334.
Distrutti E, Azpiroz F, Soldevilla A, Malagelada JR (2000) Gastric wall tension determines
perception of gastric distention. Gastroenterology; 118: 641-643.
Dou Y, Gregersen S, Zhao J, Zhuang F, Gregersen H. Effect of re-feeding after starvation on
biomechanical properties in rat small intestine. Med. Eng. Phys. 2001; 23: 557-566.
Dou Y, Gregersen S, Zhao J, Zhuang F, Gregersen H. Morphometric and biomechanical
intestinal remodeling induced by fasting in rats. Dig. Dis. Sci. 2002a; 47: 1158-1168.
Dou Y, Lu X, Zhao J, Gregersen H. Morphometric and biomechanical remodelling in the
intestine after small bowel resection in the rat. Neurogastroenterol Motil. 2002b; 14: 43-
53.
Dou Y, Zhao J, Gregersen H. Morphology and stress-strain properties along the small
intestine in the rat. J. Biomech. Eng. 2003; 125: 266-273.
Dou Y, Fan Y, Zhao J, Gregersen H. Longitudinal residual strain and stress-strain relationship
in rat small intestine. Biomed. Eng. Online. 2006 Jun 7;5:37.
Drewes AM, Petersen P, Rössel P, Gao C, Hansen JB, Arendt-Nielsen L. Sensitivity and
distensibility of the rectum and sigmoid colon in patients with irritable bowel syndrome.
Scand. J. Gastroenterol. 2001; 36: 827-832.
Drewes AM, Pedersen J, Liu W, Arendt-Nielsen L, Gregersen H. Controlled mechanical
distension of the human oesophagus: sensory and biomechanical findings. Scand. J.
Gastroenterol. 2003a; 38: 27-35.
Drewes AM, Gregersen H. Multimodal pain stimulation of the gastrointestinal tract. World J.
Gastroenterol. 2006; 12: 2477-2486.
Duch BU, Petersen JA, Vinter-Jensen L, Gregersen H. Elastic properties in the
circumferential direction in isolated rat small intestine. Acta. Physiol. Scand. 1996,157:
157-163.
Dvorkin LS, Chan CL, Knowles CH, Williams NS, Lunniss PJ, Scott SM. Anal sphincter
morphology in patients with full-thickness rectal prolapse. Dis. Colon. Rectum. 2004;
47:198-203.
Egorov VI, Schastlivtsev IV, Prut EV, Baranov AO, Turusov RA. Mechanical properties of
the human GI tract. J. Biomech. 2002; 35: 1417-1425.
Elbrønd H, Tøttrup A, Forman A. Mechanical properties of isolated smooth muscle from
rabbit sphincter of Oddi and duodenum. Scand. J. Gastroenterol. 1991; 26: 289-294.
Fackler K, Klein L, Hiltner A. Polarizing light microscopy of intestine and its relationship to
mechanical behaviour. J. Microsc. 1981; 124(Pt 3): 305-311.
Fan Y, Gregersen H, Kassab GS. A two-layered mechanical model of the rat esophagus.
Experiment and theory. Biomed Eng Online. 2004 Nov 1; 3(1): 40.
Fan Y, Zhao J, Liao D, Gregersen H. The effect of digestion of collagen and elastin on
histomorphometry and the zero-stress state in rat esophagus. Dig. Dis. Sci. 2005; 50:
1497-1505.
Folwaczny C, Riepl R, Tschop M, Landgraf R. GI involvement in patients with diabetes
mellitus: Part I (first of two parts). Epidemiology, pathophysiology, clinical findings.
Z.Gastroenterol. 1999: 37: 803-815.
Ford MJ, Camilleri M, Wiste JA, Hanson RB. Differences in colonic tone and phasic
response to a meal in the transverse and sigmoid human colon. Gut. 1995; 37: 264-269.
Frokjaer JB, Liao D, Bergmann A, McMahon BP, Steffensen E, Drewes AM, Gregersen H.
Three-dimensional biomechanical properties of the human rectum evaluated with
magnetic resonance imaging. Neurogastroenterol Motil. 2005; 17: 531-540.
Frokjaer JB, Andersen SD, Lundbye-Christensen S, Funch-Jensen P, Drewes AM, Gregersen
H. Sensation and distribution of stress and deformation in the human oesophagus.
Neurogastroenterol Motil. 2006a; 18: 104-114.
Frokjaer JB, Andersen SD, Drewes AM, Gregersen H. Ultrasound-determined geometric and
biomechanical properties of the human duodenum. Dig. Dis. Sci. 2006b; 51: 1662-1669.
Frokjaer JB, Liao D, Steffensen E, Dimcevski G, Bergmann A, Drewes AM, Gregersen H.
Geometric and mechanosensory properties of the sigmoid colon evaluated with magnetic
resonance imaging. Neurogastroenterol Motil. 2007a; 19: 253-262.
Frokjaer JB, Andersen SD, Ejskjaer N, Funch-Jensen P, Drewes AM, Gregersen H. Impaired
contractility and remodeling of the upper GI tract in diabetes mellitus type-1. World J.
Gastroenterol. 2007b; 13 (36): 4881-4890.
Fung YC. What are the residual stresses doing in our blood vessels? Ann. Biomed. Eng.
1991:19: 237-249.
Fung YC, Liu SQ. Strain distribution in small blood vessels with zero-stress state taken into
consideration. Am. J. Physiol. 1992: 262: H544-H552.
Fung YC. Biomechanics. Properties of living tissues. Berlin, Springer-Verlag,1993.
Gabella G, Blundell D. Gap junctions of the muscles of the small and large intestine. Cell
Tissue Res. 1981: 219: 469-488.
Gao C, Zhao J, Gregersen H. Histomorphometry and strain distribution in pig duodenum with
reference to zero-stress state. Dig. Dis. Sci. 2000a; 45: 1500-1508.
Gao C, Gregersen H. Biomechanical and morphological properties in rat large intestine. J.
Biomech. 2000b; 33:1089-1097.
Gao C, Arendt-Nielsen L, Liu W, Petersen P, Drewes AM, Gregersen H. Sensory and
biomechanical responses to ramp-controlled distension of the human duodenum. Am. J.
Physiol Gastrointest Liver Physiol. 2003; 284: G461-G471.
Gao F, Liao D, Zhao J, Drewes AM, Gregersen H. Numerical analysis of pouch filling and
emptying after laparoscopic gastric banding surgery. Obes. Surg. 2008; 18: 243-250.
Gilja OH, Lunding J, Hausken T, Gregersen H. Gastric accommodation assessed by
ultrasonography. World J. Gastroenterol. 2006; 12: 2825-2829.
Gilja OH, Hatlebakk JG, Odegaard S, Berstad A, Viola I, Giertsen C, Hausken T, Gregersen
H. Advanced imaging and visualization in GI disorders. World J. Gastroenterol. 2007;
13: 1408-1421.
Glavind EB, Forman A, Madsen G, Svane D, Andersson KE, Tøttrup A. Mechanical
properties of isolated smooth muscle from human rectum and internal anal sphincter. Am.
J. Physiol. 1993; 265(4 Pt 1): G792-G798.
Gravesen FH, McMahon BP, Drewes AM, Gregersen H. Measurement of the axial force
during primary peristalsis in the oesophagus using a novel electrical impedance
technology. Physiol. Meas. 2008; 29: 389-399.
Gregersen H, Jensen LS, Djurhuus JC. Changes in oesophageal wall biomechanics after
portal vein banding and variceal sclerotherapy measured by a new technique. An
experimental study in rabbits. Gut. 1988; 29: 1699-1704.
Gregersen H, Orvar K, Christensen J. Biomechanical properties of duodenal wall and
duodenal tone during phase I and phase II of the MMC. Am. J. Physiol. 1992a; 263(5 Pt
1): G795-G801.
Gregersen H, Giversen IM, Rasmussen LM, Tøttrup A. Biomechanical wall properties and
collagen content in the partially obstructed opossum esophagus. Gastroenterology.
1992b; 103: 1547-1551.
Gregersen H, Jorgensen CS, Dall FH, Jensen SL Characteristics of spontaneous and evoked
motility in the isolated perfused porcine duodenum. J. Appl. Physiol. 1992c; 73: 9-19.
Gregersen H, Dall FH, Jorgensen CS, Jensen SL, Ahren B. Effects of noradrenaline and
galanin on duodenal motility in the isolated perfused porcine pancreatico-duodenal block.
Regul Pept. 1992d; 39: 157-167.
Gregersen H, Kassab G. Biomechanics of the GI tract. Neurogastroenterol. Motil. 1996; 8:
277-297.
Gregersen H, Kassab G, Pallencaoe E, Lee C, Chien S, Skalak R, Fung YC. Morphometry
and strain distribution in guinea pig duodenum with reference to the zero-stress state. Am.
J. Physiol. 1997; 273(4 Pt 1): G865-G874.
Gregersen H, Lee TC, Chien S, Skalak R, Fung YC. Strain distribution in the layered wall of
the esophagus. J. Biomech. Eng. 1999; 121: 442-448.
Gregersen H, Kassab GS, Fung YC. The zero-stress state of the GI tract: biomechanical and
functional implications. Dig. Dis. Sci. 2000; 45: 2271-2281.
Gregersen H, Weis SM, McCulloch AD. Oesophageal morphometry and residual strain in a
mouse model of osteogenesis imperfecta. Neurogastroenterol Motil. 2001; 13: 457-464.
Gregersen H. Biomechanics of the Gastrointestinal Tract. London : Springer-Verlag, 2002a:
1-268.
Gregersen H, Lundby L, Overgaard J. Early and late effects of irradiation on morphometry
and residual strain of mouse rectum. Dig. Dis. Sci. 2002b; 47: 1472-1479.
Gregersen H, Gilja OH, Hausken T, Heimdal A, Gao C, Matre K, Ødegaard S, Berstad A.
Mechanical properties in the human gastric antrum using B-mode ultrasonography and
antral distension. Am. J. Physiol Gastrointest Liver Physiol. 2002c; 283: G368-G375.
Gregersen H, Lu X, Zhao J. Physiological growth is associated with esophageal
morphometric and biomechanical changes in rats. Neurogastroenterol Motil. 2004; 16:
403-412.
Gregersen H, Hausken T, Yang J, Ødegaard S, Gilja OH. Mechanosensory properties in the
human gastric antrum evaluated using B-mode ultrasonography during volume-controlled
antral distension. Am. J. Physiol Gastrointest Liver Physiol. 2006b; 290: G876-G882.
Gregersen H, Liao D, Fung YC. Determination of homeostatic elastic moduli in two layers of
the esophagus. J. Biomech. Eng. 2008; 130: 011005.
Haffner JFW, Stadaas J. Pressure responses to cholinergic and adrenergic agents in the
fundus, corpus, and antrum of isolated rabbit stomachs. Acta. Chir. Scand. 1972; 138:
713-719.
He CL, Soffer EE, Ferris CD, Walsh RM, Szurszewski JH, Farrugia G. Loss of interstitial
cells of cajal and inhibitory innervation in insulin-dependent diabetes. Gastroenterology.
2001; 121: 427-434.
Hecker L, Baar K, Dennis RG, Bitar KN. Development of a three-dimensional physiological
model of the internal anal sphincter bioengineered in vitro from isolated smooth muscle
cells. Am. J. Physiol Gastrointest Liver Physiol. 2005; 289: G188-G196.
Helm JF, Dodds WJ, Christensen J, Sarna SK. Control mechanism of spontaneous in vitro
contractions of the opossum sphincter of Oddi. Am. J. Physiol. 1985; 249(5 Pt 1): G572-
G579.
Helm JF, Dodds WJ, Christensen J, Sarna SK. Intramural neural control of opossum sphincter
of Oddi. Am. J. Physiol. 1989; 257(6 Pt 1): G925-G929.
Hind JA, Nicosia MA, Roecker EB, Carnes ML, Robbins J. Comparison of effortful and
noneffortful swallows in healthy middle-aged and older adults. Arch. Phys. Med. Rehabil.
2001; 82: 1661-1665.
Holland CT, Satchell PM, Farrow BR. Oesophageal compliance in naturally occurring canine
megaoesophagus. Aust. Vet. J. 1993; 70: 414-420.
Horowitz M, Edelbroek M, Fraser R, Maddox A, Wishart J. Disordered gastric motor
function in diabetes melliitus. Scand. J. Gastroenterol. 1991; 26: 673-684.
Hunter PJ, Nielsen PM, Bullivant D. The IUPS Physiome Project. International Union of
Physiological Sciences. Novartis Found Symp. 2002; 247: 207-217.
Hunter PJ, Nielsen PM. A strategy for integrative computational physiology. Physiology
(Bethesda). 2005; 20: 316-325.
Indireshkumar K, BrassEur. J.G, Faas H, Hebbard GS, Kunz P, Dent J, Feinle C, Li M,
Boesiger P, Fried M, Schwizer W. Relative contributions of "pressure pump" and
"peristaltic pump" to gastric emptying. Am. J. Physiol Gastrointest Liver Physiol. 2000;
278: G604-G616.
Jacob P, Kahrilas PJ, Logemann JA, Shah V, Ha T. Upper esophageal sphincter opening and
modulation during swallowing. Gastroenterology. 1989; 97: 1469-1478.
Jeays AD, Lawford PV, Gillott R, Spencer P, Barber DC, Bardhan KD, Hose DR.
Characterisation of the haemodynamics of the superior mesenteric artery. J. Biomech.
2007; 40: 1916-1926.
Johnston BT, Li Q, Castell JA, Castell DO. Swallowing and esophageal function in
Parkinson's disease. Am. J. Gastroenterol. 1995; 90: 1741-1746.
Jorgensen CS, Dall FH, Storkholm J, Jensen SL, Gregersen H. Elastic properties of the
isolated perfused porcine duodenum. Dig. Dis.. 1991: 9: 401-407.
Jorgensen CS, Dall FH, Jensen SL, Gregersen H. A new combined high-frequency
ultrasound-impedance planimetry measuring system for the quantification of organ wall
biomechanics in vivo. J. Biomech. 1995; 28: 863-867.
Juhl CO, Vinter-Jensen L, Djurhuus JC, Gregersen H, Dajani EZ. Biomechanical properties
of the oesophagus damaged by endoscopic sclerotherapy. An impedance planimetric
study in minipigs. Scand. J. Gastroenterol. 1994; 29: 867-873.
Jørgensen CS, Ahrensberg JM, Gregersen H, Flyvberg A. Tension-strain relations and
morphometry of rat small intestine in experimental diabetes. Dig. Dis. Sci. 2001; 46: 960-
967.
Karakida T, Homma S. Compliance changes of the GI tract in streptozotocin-induced diabetic
rats. Jpn J. Physiol. 1989: 39: 559-570.
Keinke O, Schemann M, Ehrlein HJ. Mechanical factors regulating gastric emptying of
viscous nutrient meals in dogs. Q. J. Exp. Physiol. 1984; 69: 781-795.
Kahrilas PJ, Logemann JA, Krugler C, Flanagan E. Volitional augmentation of upper
esophageal sphincter opening during swallowing. Am. J. Physiol. 1991; 260: G450-G456
Karakida T, Homma S. Compliance changes of the GI tract in streptozotocin-induced diabetic
rats. Jpn J. physiol. 1989; 39: 559-570.
Kissmeyer-Nielsen P, Christensen H, Laurberg S. Decrease in collagenous proteins and
mechanical strength of distal colon after diverting colostomy in rats. Int. J. Colorectal
Dis. 1993; 8: 120-124.
Kumar A, Attaluri A, Hashmi S, Schulze KS, Rao SS. Visceral hypersensitivity and impaired
accommodation in refractory diabetic gastroparesis. Neurogastroenterol Motil. 2008; 20:
635-642.
Kuriyama H, Mishima K, Suzuki H. Some differences in contractile responses of isolated
longitudinal and circular muscle from the guinea-pig stomach. J. Physiol. 1975; 251:
317-331.
Li J, Zhao J, Liao D, Gregersen H. Effect of smooth muscle tone on morphometry and
residual strain in rat duodenum, jejunum and ileum. J. Biomech. 2008; 41: 2667-2672.
Li L, Zhang J, Lu G. [Tension-stress effect on genesis and growth of the external anal
sphincter] [Article in Chinese]. Zhonghua Wai Ke Za Zhi. 1995; 33: 123-126.
Li M, BrassEur. J.G, Dodds WJ. Analyses of normal and abnormal esophageal transport
using computer simulations. Am. J. Physiol. 1994; 266: G525-G543.
Liao D, Fan Y, Zeng Y, Gregersen H. Stress distribution in the layered wall of the rat
oesophagus. Med. Eng. Phys. 2003a; 25: 731-738.
Liao D, Yang J, Zhao J, Zeng Y, Vinter-Jensen L, Gregersen H. The effect of epidermal

growth factor on the incremental Young's moduli in the rat small intestine. Med. Eng.
Phys. 2003b; 25: 413-418.
Liao D, Gregersen H, Hausken T, Gilja OH, Mundt M, Kassab G. Analysis of surface
geometry of the human stomach using real-time 3-D ultrasonography in vivo.
Neurogastroenterol Motil. 2004a; 16: 315-324.
Liao D, Zhao J, Fan Y, Gregersen H. Two-layered quasi-3D finite element model of the
oesophagus. Med. Eng. Phys. 2004b; 26: 535-543.
Liao D, Zhao J, Gregersen H. Regional surface geometry of the rat stomach based on three-
dimensional curvature analysis. Phys Med. Biol. 2005; 50: 231-246.
Liao D, Zhao J, Gregersen H. Three-dimensional geometry analysis of the stomach in type II
diabetic GK rats. Diabetes Res. Clin. Pract. 2006a; 71: 1-13.
Liao D, Cassin J, Zhao J, Gregersen H.The geometric configuration and morphometry of the
rabbit oesophagus during luminal pressure loading. Physiol. Meas. 2006b; 27: 703-711.
Liao D, Zhao J, Yang J, Gregersen H. The oesophageal zero-stress state and mucosal folding
from a GIOME perspective. World J. Gastroenterol. 2007; 13: 1347-1351.
Liao D, Lelic D, Gao F, Drewes AM, Gregersen H. Biomechanical functional and sensory
modelling of the GI tract. Philos. Transact. A Math. Phys. Eng. Sci. 2008; 366: 3281-
3299.
Liao DH, Zhao JB, Gregersen H. Gastrointestinal tract modelling in health and disease. World
J. Gastroenterol. 2009; 15: 169-176.
Lin AS, Buist ML, Cheng LK, Smith NP, Pullan AJ. Computational simulations of the human
magneto- and electroenterogram. Ann. Biomed. Eng, 2006a; 34: 1322-1331.
Lin AS, Buist ML, Smith NP, Pullan AJ. Modelling slow wave activity in the small intestine.
J. Theor. Biol. 2006b; 242: 356-362.
Loening-Baucke V, Metcalf AM, Shirazi S. Anorectal manometry in active and quiescent
ulcerative colitis. American Journal of Gastroenterology. 1989; 84: 892-897.
Logemann JA, Pauloski BR, Rademaker AW, Colangelo LA, Kahrilas PJ, Smith CH.
Temporal and biomechanical characteristics of oropharyngeal swallow in younger and
older men. J. Speech Lang Hear Res. 2000; 43: 1264-1274.
Lord RV, Demeester SR, Peters JH, Hagen JA, Elyssnia D, Sheth CT, Demeester TR. iatal
Hernia, Lower Esophageal Sphincter Incompetence, and Effectiveness of Nissen
Fundoplication in the Spectrum of Gastroesophageal Reflux Disease. J. Gastrointest
Surg. 2008 Dec 3. [Epub ahead of print].
Lu X, Gregersen H. Regional distribution of axial strain and circumferential residual strain in
the layered rabbit oesophagus. J. Biomech. 2001; 34: 225-233.
Lu X, Zhao J, Gregersen H. Small intestinal morphometric and biomechanical changes during
physiological growth in rats. J. Biomech. 2005; 38: 417-426.
Malagelada JR, Rees WD, Mazzotta LJ, Go VL.Gastric motor abnormalities in diabetic and
postvagotomy gastroparesis: effect of metoclopramide and bethanechol.
Gastroenterology. 1980; 78: 286-293.
Mandrek K, Kreis S. Regional differentiation of gastric and of pyloric smooth muscle in the
pig: mechanical responses to acetylcholine, histamine, substance P, noradrenaline and
adrenaline. J. Auton. Pharmacol. 1992; 12: 37-49.
Massey BT, Simuncak C, LeCapitaine-Dana NJ, Pudur S. Transient lower esophageal

sphincter relaxations do not result from passive opening of the cardia by gastric
distention. Gastroenterology. 2006; 130: 89-95.
Matsufuji H, Yokoyama J. Neural control of the internal anal sphincter motility. J. Smooth
Muscle Res. 2003; 39: 11-20.
McMahon BP, Frokjaer JB, Kunwald P, Liao D, Funch-Jensen P, Drewes AM, Gregersen H.
The functional lumen imaging probe (FLIP) for evaluation of the esophagogastric
junction. Am. J. Physiol Gastrointest Liver Physiol. 2007a; 292: G377-G384.
McMahon BP, Odie KD, Moloney KW, Gregersen H. Computation of flow through the
oesophagogastric junction.World J. Gastroenterol. 2007b; 13: 1360-1364.
Mearin F, Camilleri M, Malagelada JR. Pyloric dysfunction in diabetics with recurrent nausea
and vomiting. Gastroenterology. 1986; 90: 1919-1925.
Mearin F, Papo M, Malagelada JR. Impaired gastric relaxation in patients with achalasia. Gut.
1995; 36: 363-368.
Mearin F, Fonollosa V, Vilardell M, Malagelada JR. Mechanical properties of the gastro-
esophageal junction in health, achalasia, and scleroderma. Scand. J. Gastroenterol. 2000;
35: 705-710.
Milenov K, Golenhofen K. Contractile responses of longitudinal and circular smooth muscle
of the canine stomach to prosglanding E and F2 alpha. Prostaglandings Leukotrienes
Med. 1982; 8: 287-300.
Miller LS, Dai Q, Sweitzer BA, Thangada V, Kim JK, Thomas B, Parkman H, Soliman AM.
Evaluation of the upper esophageal sphincter (UES) using simultaneous high-resolution
endoluminal sonography (HRES) and manometry. Dig. Dis. Sci. 2004; 49: 703-709.
Moragas G, Azpiroz F, Pavia J, Malagelada JR (1993) Relations among intragastric pressure,
postcibal perception, and gastric emptying. Am. J. Physiol. 264: G1112-G1117.
Natalia A, Carniel EL, Pavan PG, Dario P, Izzo I, Menciassi A. Evaluation of the constitutive
parameters for the numerical analysis of non linear mechanical response of soft tissues.
Proceedings of Computer Methods in Biomechanics and Biomedical Engineering 2006a,
Antibes, France.
Natalia A, Carniel EL, Pavan PG, Dario P, Izzo I, Menciassi A. Hyperelastic models for the
analysis of soft tissue mechanics: definition of constitutive parameters. IEEE/RAS-EMBS
International Conference on Biomedical Robotics and Biomechatronics. 2006b (pp. 1-8).
Nicosia MA, BrassEur. J.G, Liu JB, Miller LS. Local longitudinal muscle shortening of the
human esophagus from high-frequency ultrasonography. Am. J. Physiol. Gastrointest.
Liver Physiol. 2001; 281: G1022-G1033.
Noda T, Iwakiri R, Fujimoto K, Yoshida T, Utsumi H, Sakata H, Hisatomi A, Aw TY.
Suppression of apoptosis is responsible for increased thickness of intestinal mucosa in
streptozotocin-induced diabetic rats. Metabolism. 2001; 50: 259-264.
Noriyuki O, Sengupta JN, Gebhart GF. Mechanosensitive properties of gastric vagal afferent
Fibers in the rat. J. Neurophysiol. 1999; 82: 2210–2220.
Orvar KB, Gregersen H, Christensen J. Biomechanical characteristics of the human
esophagus. Dig. Dis. Sci. 1993; 38: 197-205.
Pal A, BrassEur. J.G, Abrahamsson B. A stomach road or "Magenstrasse" for gastric
emptying. J. Biomech. 2007; 40: 1202-1210.
Pandolfino JE, Shi G, Curry J, Joehl RJ, BrassEur. J.G, Kahrilas PJ. Esophagogastric junction
distensibility: a factor contributing to sphincter incompetence. Am. J. Physiol
Park JH, Baek YH, Park DI, Kim HJ, Cho YK, Sohn CI, Jeon WK, Kim BI, Rhee PL.
Analysis of rectal dynamic and static compliances in patients with irritable bowel
syndrome. Int. J. Colorectal Dis. 2008; 23: 659-664.
Penninckx F, Lestar B, Kerremans R. The internal anal sphincter: mechanisms of control and
its role in maintaining anal continence. Baillieres Clin Gastroenterol. 1992; 6: 193-214.
Patel RS, Rao SS. Biomechanical and sensory parameters of the human esophagus at four
levels. Am. J. Physiol. 1998; 275(2 Pt 1): G187-G191.
Petersen P, Gao C, Rössel P, Qvist P, Arendt-Nielsen L, Gregersen H, Drewes AM. Sensory
and biomechanical responses to distension of the normal human rectum and sigmoid
colon. Digestion. 2001; 64: 191-199.
Pedersen J, Gao C, Egekvist H, Bjerring P, Arendt-Nielsen L, Gregersen H, Drewes AM.
Pain and biomechanical responses to distention of the duodenum in patients with
systemic sclerosis. Gastroenterology. 2003; 124: 1230-1239.
Pullan A, Cheng L, Yassi R, Buist M. Modelling gastrointestinal bioelectric activity. Prog.
Biophys. Mol. Biol. 2004; 85: 523-550.
Qiao Y, Pan E, Chakravarthula SS, Han F, Liang J, Gudlavalleti S. Measurement of
mechanical properties of rectal wall. J. Mater Sci. Mater Med. 2005; 16: 183-188.
Rao SS, Read NW, Davison PA, Bannister JJ, Holdsworth CD. Anorectal sensitivity and
responses to rectal distention in patients with ulcerative colitis. Gastroenterology. 1987;
93, 1270-1275.
Rao SS, Hayek B, Summers RW. Impedance planimetry: an integrated approach for assessing
sensory, active, and passive biomechanical properties of the human esophagus. Am. J.
Gastroenterol. 1995; 90: 431-438.
Rao SS, Gregersen H, Hayek B, Summers RW, Christensen J. Unexplained chest pain: the
hypersensitive, hyperreactive, and poorly compliant esophagus. Ann Intern Med. 1996;
124: 950-958.
Rao SS, Mudipalli RS, Mujica VR, Patel RS, Zimmerman B. Effects of gender and age on
esophageal biomechanical properties and sensation. Am. J. Gastroenterol. 2003; 98:
1688-1695.
Rao GN, Drew PJ, Monson JR, Duthie GS. Incremental elastic modulus--a challenge to
compliance. Int. J. Colorectal Dis. 1997; 12: 33-36.
Rayner CK, Verhagen MAMT, Hebbard GS, DiMatteo AC, Doran SM, Horowitz M.
Proximal gastric compliance and perception of distension in type 1 diabetes mellitus:
Effect of hyperglycemia. Am. J. Gastroenterology. 2000; 95: 1175-1183.
Reddy H, Staahl C, Arendt-Nielsen L, Gregersen H, Drewes AM, Funch-Jensen P. Sensory
and biomechanical properties of the esophagus in non-erosive reflux disease. Scand. J.
Gastroenterol. 2007; 42: 432-440.
Remes-Troche JM, Maher J, Mudipalli R, Rao SS. Altered esophageal sensory-motor
function in patients with persistent symptoms after Nissen fundoplication. Am. J. Surg.
2007; 193: 200-205.
Rothstein RD. GI motility disorders in diabetes mellitus. Am. J. Gastroenterology. 1990; 85:
782-785.
Salet GA, Samsom M, Roelofs JM, van Berge Henegouwen GP, Smout AJ, Akkermans LM.
Responses to gastric distension in functional dyspepsia. Gut. 1998; 42: 823-829.
Schulze-Delrieu K, Shirazi SS. Pressure and length adaptations in isolated cat stomach. Am. J.
Physiol 1987; 252: G92-G99.
Schulze-Delrieu K. Intrinsic differences in the filling response of the guinea pig duodenum
and ileum. J. Lab. Clin. Med. 1991: 117: 44-50.
Schulze-Delrieu K, Herman RJ, Shirazi SS, Brown BP. Contractions move contents by
changing the configuration of the isolated cat stomach. Am. J. Physiol 1998; 274: G359-
G369.
Schwizer WA, Steingotter M, Fox T, Zur T, Thumshirn M, Bosiger P, Fried M. Non-invasive
measurement of gastric accommodation in humans. Gut. 2002; 51(Suppel): i59-i62.
Secondulfo M, Iafusco D, Carratu R, deMagistris L, Sapone A, Generoso M, Mezzogiomo A,
Sasso FC, Carteni M, De Rosa R, Prisco F, Esposito V. Ultrastructural mucosal
alterations and increased intestinal permeability in non-celiac, type I diabetic patients.
Dig. Liver Dis. 2004; 36: 35-45.
Shafik A. A concept of the anatomy of the anal sphincter mechanism and the physiology of
defecation. Dis. Colon. Rectum. 1987; 30: 970-982.
Shafik A. Effect of duodenal distension on the pyloric sphincter and antrum and the gastric
corpus: duodenopyloric reflex. World J. Surg. 1998; 22: 1061-1064.
Shirazi S, Schulze-Delrieu K, Brown CK. Duodenal resistance to the emptying of various
solutions from the isolated cat stomach. J. Lab. Clin. Med. 1988; 111: 654-660.
Siegle ML, Ehrlein HJ. Interdigestive contractile patterns of the ileum in dogs. Am. J. Physiol.
1987: 253: G452-G460.
Siegle ML, Ehrlein HJ. Neurotensin changes the motor pattern in canine ileum from
propulsive to segmenting. Dig. Dis. Sci. 1989: 34: 1521-1527.
Siegle ML, Buhner S, Schemann M, Schmid HR, Ehrlein HJ. Propagation velocities and
frequencies of contractions along canine small intestine. Am. J. Physiol. 1990: 258:
G738-G744.
Smith JB, Zhao JB, Dou YL, Gregersen H. Time-dependent viscoelastic properties along rat
small intestine. World J. Gastroenterol. 2005; 11: 4974-4878.
Storkholm JH, Villadsen GE, Jensen SL, Gregersen H. Passive elastic wall properties in
isolated guinea pig small intestine. Dig. Dis. Sci. 1995: 40: 976-982.
Storkholm JH, Villadsen GE, Jensen SL, Gregersen H. Mechanical properties and collagen
content differ between isolated guinea pig duodenum, jejunum, and distal ileum. Dig.
Dis. Sci. 1998; 43: 2034-2041.
Storkholm JH, Zhao J, Villadsen GE, Hager H, Jensen SL, Gregersen H. Biomechanical
remodeling of the chronically obstructed Guinea pig small intestine. Dig. Dis. Sci. 2007;
52: 336-346.
Storkholm JH, Zhao J, Villadsen GE, Gregersen H. Spontaneous and bolus-induced motility
in the chronically obstructed guinea-pig small intestine in vitro. Dig. Dis. Sci. 2008; 53:
413-420.
Sun WM, Hebbard GS, Malbert CH, Jones KL, Doran S, Horowitz M, Dent J. Spatial
patterns of fasting and fed antropyloric pressure waves in humans. J. Physiol. 1997; 503:
455-462.
Szurszewsky JH. A 100-year perspective on gastrointestinal motility. Am. J. Physiol. 1998;
274: G447-G453.
Takeda T, Kassab G, Liu J, Puckett JL, Mittal RR, Mittal RK. A novel ultrasound technique
to study the biomechanics of the human esophagus in vivo. Am. J. Physiol. Gastrointest
Liver Physiol. 2002; 282: G785-G793.
Takeda T, Kassab G, Liu J, Nabae T, Mittal RK. Effect of atropine on the biomechanical
properties of the oesophageal wall in humans. J. Physiol. 2003; 547(Pt 2): 621-628.
Tormo MA, Martinez IM, Romero dT, Gil-Exojo I, Campillo JE. Morphological and
enzymatic changes of the small intestine in an n0-STZ diabetes rat model. Exp. Clin.
Endocrinol Diabetes. 2002; 110: 119-123.
Vanags I, Petersons A, Ose V, Ozolanta I, Kasyanov V, Laizans J, Vjaters E, Gardovskis J,
Vanags A. Biomechanical properties of oesophagus wall under loading. J. Biomech.
2003; 36: 1387-1390.
Verne GN, Sninsky CA. Diabetes and the GI tract. Gastroenterol Clin. North Am. 1998: 27:
861-874.
Watters DA, Smith AN, Eastwood MA, Anderson KC, Elton RA. Mechanical properties of
the rat colon: the effect of age, sex and different conditions of storage. Q. J. Exp. Physiol.
1985a; 70: 151-162.
Watters DA, Smith AN, Eastwood MA, Anderson KC, Elton RA, Mugerwa JW. Mechanical
properties of the colon: comparison of the features of the African and European colon in
vitro. Gut. 1985b; 26: 384-392.
Whitehead WE, Delvaux M, Working Team. Standardization of barostat procedures for
testing smooth tone and sensory thresholds in the GI tract. Dig. Dis. Sci. 1997; 42: 223-
241.
Williams RB, Pal A, BrassEur. J.G, Cook IJ. Space-time pressure structure of pharyngo-
esophageal segment during swallowing. Am. J. Physiol. Gastrointest Liver Physiol. 2001;
28: G1290-G1300.
Williams RB, Wallace KL, Ali GN, Cook IJ. Biomechanics of failed deglutitive upper
esophageal sphincter relaxation in neurogenic dysphagia. Am. J. Physiol. Gastrointest
Liver Physiol. 2002; 283: G16-G26.
Yang J, Zhao JB, Zeng YJ, Gregersen H. Biomechanical properties of ileum after systemic
treatment with epithelial growth factor. World J. Gastroenterol. 2003a; 9: 2278-2283.
Yang J, Zhao J, Zeng Y, Vinter-Jensen L, Gregersen H. Morphological properties of zero-
stress state in rat large intestine during systemic EGF treatment. Dig. Dis. Sci. 2003b; 48:
442-448.
Yang J, Liao D, Zhao J, Gregersen H. Shear modulus of elasticity of the esophagus. Ann.
Biomed. Eng. 2004a; 32: 1223-1230.
Yang J, Zhao J, Zeng Y, Gregersen H. Biomechanical properties of the rat oesophagus in
experimental type-1 diabetes. Neurogastroenterol Motil. 2004b; 16: 195-203.
Yang J, Zhao J, Liao D, Gregersen H. Biomechanical properties of the layered oesophagus
and its remodelling in experimental type-1 diabetes. J. Biomech. 2006; 39: 894-904.
Yang J, Zhao J, Nakaguchi T, Gregersen H. Biomechanical changes in oxazolone-induced
colitis in BALB/C mice. J Biomech. 2009; 42: 811-817.
Yang W, Fung TC, Chian KS, Chong CK. Directional, regional, and layer variations of
mechanical properties of esophageal tissue and its interpretation using a structure-based
constitutive model. J. Biomech. Eng. 2006; 128: 409-418.
Yang W, Fung TC, Chian KS, Chong CK. Three-dimensional finite element model of the
two-layered oesophagus, including the effects of residual strains and buckling of mucosa.
Proc. Inst. Mech. Eng [H]. 2007a; 221: 417-426.
Yang W, Fung TC, Chian KS, Chong CK. Instability of the two-layered thick-walled
esophageal model under the external pressure and circular outer boundary condition. J.
Biomech. 2007b; 40: 481-490.
Yang W, Fung TC, Chian KS, Chong CK. Finite element simulation of food transport
through the esophageal body. World J. Gastroenterol. 2007c; 13: 1352-1359.
Yassi R, Cheng LK, Al-Ali S, Smith NP, Pullan AJ, Windsor JA. An anatomically based
mathematical model of the gastroesophageal junction. Conf. Proc. IEEE Eng. Med. Biol.
Soc., 2004; 1: 635-638.
Zaninotto G, DeMeester TR, Schwizer W, Johansson KE, Cheng SC. The lower esophageal
sphincter in health and disease. Am. J. Surg. 1988; 155: 104-111.
Zhao JB, Sha H, Zhuang FY, Gregersen H. Morphological properties and residual strain
along the small intestine in rats. World J. Gastroenterol.2002a; 8: 312-317.
Zhao J, Sha H, Zhou S, Tong X, Zhuang FY, Gregersen H. Remodelling of zero-stress state
of small intestine in streptozotocin-induced diabetic rats. Effect of gliclazide. Dig. Liver
Dis. 2002b; 34: 707-716.
Zhao J, Yang J, Vinter-Jensen L, Zhuang F, Gregersen H. The morphometry and
biomechanical properties of the rat small intestine after systemic treatment with
epidermal growth factor. Biorheology. 2002c; 39: 719-733.
Zhao J, Yang J, Gregersen H. Biomechanical and morphometric intestinal remodelling during
experimental diabetes in rats. Diabetologia. 2003a; 46: 1688-1697.
Zhao J, Liao D, Yang J, Gregersen H. Viscoelastic behavior of small intestine in
streptozotocin-induced diabetic rats. Dig. Dis. Sci. 2003b; 48: 2271-2277.
Zhao J, Yang J, Vinter-Jensen L, Zhuang F, Gregersen H. Biomechanical properties of
esophagus during systemic treatment with epidermal growth factor in rats. Ann. Biomed.
Eng. 2003c; 31: 700-709.
Zhao J, Liao D, Gregersen H. Tension and stress in the rat and rabbit stomach are location-
and direction-dependent. Neurogastroenterol Motil. 2005; 17: 388-398.
Zhao J, Frokjaer JB, Drewes AM, Ejskjaer N. Upper GI sensory-motor dysfunction in
diabetes mellitus. World J. Gastroenterol. 2006; 12: 2846-2857.
Zhao J, Jørgensen CS, Liao D, Gregersen H. Dimensions and circumferential stress-strain
relation in the porcine esophagus in vitro determined by combined impedance planimetry
and high-frequency ultrasound. Dig. Dis. Sci. 2007a; 52: 1338-1344.
Zhao J, Chen X, Yang J, Liao D, Gregersen H. Opening angle and residual strain in a three-
layered model of pig oesophagus. J. Biomech. 2007b; 40: 3187-3192.
Zhao J, Liao D, Gregersen H. Biomechanical and histomorphometric esophageal remodeling
in type 2 diabetic GK rats. J. Diabetes Complications. 2007c; 21: 34-40.
Zhao J, Liao D, Chen P, Kunwald P, Gregersen H. Stomach stress and strain depend on
location, direction and the layered structure. J. Biomech. 2008a; 41: 3441-3447.
Zhao J, Nakaguchi T, Gregersen H. Biomechanical and Histomorphometric Colon
Remodelling in STZ-Induced Diabetic Rats. Dig. Dis. Sci. 2008b Nov 7. [Epub ahead of
print].
Zoubi SA, Mayhew TM, Sparrow RA. The small intestine in experimental diabetes: cellular
adaptation in crypts and villi at different longitudinal sites. Virchows Arch. 1995: 426:
501-507.
Chapter 6
ELECTROMYOGRAPHY IN THE 21ST CENTURY:

FROM VOLUNTARY SIGNALS TO MOTOR EVOKED
POTENTIALS
Petra S. Williams1,2 and Brian C. Clark1,3

1
Institute for Neuromusculoskeletal Research
2
School of Physical Therapy
3
Department of Biomedical Sciences at Ohio University, Ohio, USA
ABSTRACT
The force produced by skeletal muscle is controlled by the electrical signals being
sent from motor neurons to muscle fibers. These electrical signals, which are known as
action potentials, can be recorded as they travel along the muscle cell membrane and are
referred to as an electromyogram (EMG) signal. It has been more than a century since the
first recording of a voluntary EMG signal was reported, and today it has become a classic
technique for evaluating and recording the activation of skeletal muscles during human
movement. In recent years, the advent and development of transcranial magnetic
stimulation has re-invigorated EMG research, and it is now possible to safely and
painlessly evoke EMG signals directly from the motor cortex of conscious humans. This
chapter reviews the recording and measurement issues associated with EMG and its
respective applications. Particular attention is paid to its role in understanding the
neuromechanics of human movement.
INTRODUCTION
Skeletal muscle fibers produce force and allow for motor acts via the excitation-
contraction coupling processes (MacIntosh et al., 2006). As such, skeletal muscle fibers are
both electrical and mechanical in nature. The electrical activation of muscle fibers is under
direct control of the nervous system, specifically the α-motoneurons which serve as the final
common pathway to muscle fibers, and integrate inputs from descending pathways, spinal
cord interneuron circuits, and peripheral sensory afferents. Electromyography (EMG) is the
208 Petra S. Williams and Brian C. Clark
recording and analysis of the electrical activity from the muscle fiber membrane that initiates
the mechanical activity of the muscle (Basmajian and DeLuca, 1985). Throughout its long
and storied history EMG, as a technique, has been used for a wide-variety of purposes. For
example, EMG has been used to study the degree and timing patterns of muscle activation
associated with voluntary movement in exercise and ergonomics research, and has been used
to assess the physiological function of both the nervous and muscular systems. The former
approach commonly involves recording the EMG response to evoked contractions elicited at
different spatial locations along the corticospinal pathway which has proven useful for
clinical diagnostic applications as well as for developing a better basic understanding of the
physiological properties of the neuromuscular system.
The historical underpinnings of EMG date back more than 300-years. During the 17th
century, Italian physician Francesco Redi documented the connection between muscles and
the generation of electricity when he discovered that specialized muscles of Torpediniformes
fish (e.g., electric ray fish) were capable of producing electric discharges (Basmajian and
DeLuca, 1985). At the end of the 18th century, Italian physicist and physician Luigi Galvani
demonstrated that electricity could directly initiate muscle contraction (Galvani, 1791), and
by the end of the 19th century French physiologist Etienne-Jules Marey had recorded the
electrical activity associated with voluntary muscle contraction and introduced the term
‗electromyography‘ (Figure 1) (Marey, 1876). During the first half of the 20th century, as
technology advanced and more electrodes were developed, the ability to detect and the quality
of EMG signals gradually improved.
Figure 1. Etienne-Jules Marey around 1850 (left) and his myograph (right) (Marey, 1876). Marey was
the first individual to record the electromyogram signal.
During the latter part of the 20th century the utilization and understanding of EMG
dramatically improved in association with the development of computers and data recording
equipment that allowed for improvements in signal-to-noise ratio, signal recording and
processing and computer simulation studies. Today, in the 21st century, EMG continues to be
Electromyography in the 21st Century 209
used clinically and innovative approaches using EMG to studying physiology continue to be
developed (e.g., using transcranial magnetic brain stimulation to evoke motor potentials at the
level of the motor cortex).
Today, the myoelectric signal can be readily and easily detected from within the muscle
by an indwelling electrode (intramuscular EMG), such as a needle or fine wire, or non-
invasively through the skin overlying the muscle with surface electrodes (surface EMG).
Depending upon the type of electrode and task being performed, electrical potentials can be
recorded from single muscle fibers, that are commonly referred to as single motor unit
recordings because individual muscle fibers are under the direct control of individual motor
units, or from the summation of superimposed motor unit action potentials. Regardless of the
recording electrodes, the detected EMG signals are generally amplified, filtered and
converted from their analog form to a digital signal prior to their analysis which can be both
qualitatively and quantitatively analyzed depending on the scientific or clinical reasoning
behind the given recordings. However, for virtually all EMG methods, it is essential to
recognize that the quantitative and qualitative characteristics of the recorded signal do not
faithfully represent the original myoelectric signal (Basmajian and DeLuca, 1985; Farina et
al., 2004; Keenan et al., 2005; Keenan et al., 2006a; Keenan et al., 2006b). Accordingly, this
chapter will review the biophysical basis of the EMG signal, measurement issues of EMG and
signal processing techniques, and discuss current applications of electromyography in the 21st
century. A single chapter can not do complete justice to fully understanding the complexities
of EMG; rather, in this chapter we aim to provide an overview of the basic fundamental
principles of EMG and its many uses. To fully appreciate, utilize, employ and understand
electromyography one must understand these basic principles and applications, because as
Carlo J. DeLuca, Ph.D., a pioneer of modern EMG, eloquently stated: „Electromyography is a
seductive muse because it provides easy access to physiological processes that cause the
muscle to generate force, produce movement, and accomplish the countless functions that
allow us to interact with the world around us… To its detriment, electromyography is too
easy to use and consequently too easy to abuse‟ (De Luca, 1997).
BIOPHYSICAL BASIS, MEASUREMENT ISSUES AND EMG SIGNAL

PROCESSING
Biophysical Basis of EMG
Excitable cells, such as those of nerve and muscle, have the ability to generate a
propagating wave of depolarization (e.g., an action potential) (Mathews, 2003). During
muscle activity, both voluntary and evoked (a.k.a. electrically stimulated), there is an
electrical potential change in the surface membrane of the skeletal muscle cell that is
transmitted as an impulse across the sarcolemma to the interior of the muscle cell via a
complex system of tubules (Kandel et al., 2000; Mathews, 2003). In muscle tissue, the action
potentials are generated at the neuromuscular junctions which are located in the middle of the
fibers; thus, action potentials are propagated in both directions towards the end of the fibers
(Enoka, 2002; Mathews, 2003). In muscle cells, transmembrane potential remains relatively
stable over time based on the relative proportions of sodium, potassium, and chloride in the
extracellular and intracellular mediums (Mathews, 2003). When a neural impulse propagates
down an α -motoneuron (motor unit action potential) and arrives at the motor endplate a
depolarizing wave spreads across the muscle cell that, following an electro-mechanical delay
of ~ 25-50 msec (Cavanagh and Komi, 1979; Norman and Komi, 1979), results in force
production from muscle fibers. For both voluntary and evoked recordings, the origin of the
EMG signal is the electrical activity going across the muscle fiber membrane. Specifically,
the EMG signal comprises the extracellular waveform manifestation of the transmembrane
voltage reversal process resultant from the potentials generated from the active motor units
(Dumitru, 2000). In the case of numerous motor units being activated within the electrode‘s
detection area, the EMG signal represents the sum of these respective motor units (Figure 2)
(Basmajian and DeLuca, 1985; Farina et al., 2004; Keenan et al., 2006a).
Figure 2. Two simulated motor unit action potentials discharging asynchronously, and the resultant
EMG signal representing the sum of these respective motor units.
Recording and Measurement of the EMG Signal
The basic equipment required for modern EMG recordings consists of electrodes that
measure the change in voltage conducted through the tissues due to the flux of sodium and
potassium ions across the muscle cell membrane associated with propagating action potentials
connected with amplifiers, bandwidth filters (commonly 10-500 Hz for surface recordings
and 10-1,000 Hz for intramuscular recordings) and analog-to-digital converters or
oscilloscopes to display and measure the respective signals. There are a wide variety of
electrode choices today with each providing advantages and disadvantages. The appropriate
choice of electrodes eventually depends upon what the scientist or clinician is wishing to
measure and assess. For example, surface electrodes range in size from very large (30 cm
diameter) to small (< 4-mm diameter), while fine wire intramuscular electrodes are extremely
small (0.05-mm diameter). The smaller the diameter of the respective electrodes, the smaller
the detection area.
Table 1. Factors that influence the surface EMG
Factors That Influence the Surface EMG

Non-physiological
Anatomic Shape of the volume conductor
Thickness of the subcutaneous tissue layers
Tissue inhomogeneities
Distribution of the motor unit territories in the muscle
Size of the motor unit territories
Distribution and number of fibers in the motor unit territories
Length of the fibers
Spread of the endplates and tendon junctions within the motor units
Spread of the innervation zones and tendon regions among motor units
Presence of more than one pennation angle
Detection system Skin-electrode contact (impedance, noise)
Spatial filter for signal detection
Interelectrode distance
Electrode size and shape
Inclination of the detection system relative to muscle fiber orientation
Location of the electrodes over the muscle
Physical Conductivities of the tissues
Amount of crosstalk from nearby muscles
Physiological
Fiber membrane properties Average muscle fiber conduction velocity
Distribution of motor unit conduction velocities
Distribution of conduction velocities of the fibers within the motor units
Shape of the intracellular action potentials
Motor unit properties Number of recruited motor units
Distribution of motor unit discharge rates
Statistics and coefficient of variation for discharge rate
Motor unit synchronization
EMG, electromyogram.
Surface EMG recordings represent the activity of multiple motor units, whereas the
intramuscular recordings can represent the activity of multiple motor units, but can also be
used to detect single motor unit activity during low-force contractions where a limited
number of motor units are activated. With surface EMG recordings the amplitude of the EMG
signal is often considered a global measure of motor unit activity. However, because the
characteristics of the surface EMG signal depend on many other factors such as the
membrane properties of the muscle fibers and the timing of the motor unit action potentials
the EMG signal reflects both peripheral (muscle) and central (nervous) properties of the
neuromuscular system (Table 1) (Farina et al., 2004).
Regardless of the recording method (surface versus intramuscular) the measurement
principles of EMG are similar. EMG activity can be recorded using either a monopolar or a
bipolar recording arrangement (Figure 3). Monopolar recordings consist of a single electrode
being placed on or in the muscle of interest while a second neutral (reference/ground)
electrode being placed at an electrically quiescent site (e.g., bone). Bipolar recordings consist
of two electrodes being placed on or in the muscle of interest along with a neutral electrode.
The bipolar configuration allows for the determination of the electrical difference between the
two recording electrodes which results in detected signals that are not common between the
two electrodes to be dramatically attenuated. The bipolar configuration is utilized more as
these signals are generally more stable, although monopolar recordings are certainly
appropriate under certain conditions when measurement of the absolute magnitude of the
voltage is desired.
Figure 3. Schematic representation of a monopolar vs. bipolar electrode configuration. Monopolar

recordings consist of a single electrode being placed on or in the muscle of interest while a second
neutral (reference/ground) electrode being placed at an electrically quiescent site (e.g., bone), whereas
bipolar recordings consist of two electrodes being placed on or in the muscle of interest along with a
neutral electrode.
The bipolar configuration results in a dramatic reduction in the amplitude of the recorded
signal, and the degree of effect varies depending on the interelectrode distance (distance
between the two recording electrodes) and their relative location to the neuromuscular
junction (Merletti et al., 2001). Specifically, the amplitude will be lower with a smaller
interelectrode difference as the signals detected between the two electrodes will be more
similar due to the spatial proximity, and if the two electrodes span the neuromuscular junction
(innervation zone) the amplitude will be reduced as the electrodes will record symmetrical
potentials propagating in each direction.
EMG Signal Processing
In most applications, quantification of the EMG signal is desired. Throughout history

there have been a wide number of approaches used, but in general the basic goal of all is to
quantify a given component associated with the amplitude or frequency characteristics of the
EMG signal. First, consideration will be given to quantifying the ‗interference EMG‘ signal,
that is, the EMG signal resultant of the detection of many motor units being asynchronously
and concurrently active during a task (the voluntary EMG signal) (Figure 4).
Figure 4. Example of the interference EMG signal recorded with bipolar surface electrodes from the
vastus lateralis muscle during a maximal voluntary knee extension task (intereference EMG signal
shown in the top trace and the exerted force shown in bottom trace). Because the interference EMG
signal varies in both the positive and negative direction the mean of the signal is zero. Thus, to quantify
the amplitude of the signal mathematical processing is required. Common approaches involve
averaging the full-wave rectified EMG (2nd trace from top), calculating the root mean squared value
(3rd trace from top; over a 500-msec epoch in this example) or applying a linear envelope (4th trace
from top; using a 10-Hz low-pass Blackman-61dB filter).
Because the interference EMG signal varies in both the positive and negative direction
the mean of the signal is zero. Thus, to quantify the amplitude of the signal mathematical
processing is required. The following are brief descriptions of common methods for
quantifying the amplitude of the interference EMG signal:
 Average Rectified EMG: This quantification process involves full-wave rectifying of

the EMG signal (converting negative values to positive values), and then taking the
mean of the selected time of interest.
 Root Mean Squared EMG: This quantification process involves calculating the
square root of the mean squared values of the EMG signal at any given time point. It
is also sometimes called the quadratic mean, and in cases where the baseline is at
zero (as is desirable) this also simply represents the standard deviation. This
quantification approach does not require full-wave rectification as the calculation
incorporates the squared values of the original interference EMG signal.
 Linear Envelope EMG: The linear envelope approach involves passing a low-pass
filter through the full-wave rectified signal. As such, it is a type of moving average
indicator of EMG amplitude. Low-pass filter frequencies tend to range between 3-50
Hz, with a 10-Hz frequency being relatively common.
Regardless of the quantification method employed, interpretation of the absolute value of

EMG amplitudes between different individuals is, in general, not appropriate as there are
numerous non-physiologic influences on the EMG signal (e.g., adipose tissue dramatically
attenuates the EMG signal; Table 1) (Farina et al., 2004). To circumvent this issue, it is
necessary to normalize the EMG amplitude of a given task of interest to a reference value.
The most common normalization procedures for interpreting EMG amplitudes between
individuals or groups of individuals during exercise or ergonomic tasks is to express the
amplitude during the task to that associated with either a maximal voluntary contraction of the
given muscle, or better yet, to the EMG signal in response to supramaximal electrical
stimulation (commonly referred to as Mmax or CMAP [for compound muscle fiber action
potential]). Analyzing the interference EMG signal in the frequency domain can also yield
useful information regarding the signals characteristics. Perhaps the most common approach
to describe the frequency characteristics of the EMG signal is to quantify the mean or median
frequency of the power density spectrum (PDS) (Hagg, 1992).
In addition to the aforementioned quantification approaches, there are several other
methods that are more specific to other signal patterns. For example, EMG signals evoked by
stimulating along the corticospinal pathway (e.g., cortical stimulation, peripheral nerve
stimulation) results in the generation of a relatively synchronous motor unit response that
yields a summated, compound EMG signal (Figure 5A). Similarly, single motor unit
recordings yield one given action potential of interest (Figure 5B). As such, in these cases it is
common to quantify the peak-to-peak amplitude of the EMG signal or the peak of the
rectified signal. Additionally, the simple calculation of time or duration is commonly applied
to these signals (e.g., to determine the time interval between a single motor unit discharge to
then calculate its discharge rate or the variability in motor unit discharge rate).
Figure 5. A: Compound surface EMG signal from the soleus muscle in response to a single,
supramaximal electrical stimulus to a peripheral nerve (tibial nerve). Evoked potentials are commonly
quantified by the simple calculation of their peak to peak amplitude and/or the duration of a given
potential. B: Intramusular EMG recordings obtained from the medial head of the biceps brachii. Fine
wires were inserted into the muscle and a single motor unit was recorded discharging at a rate of 6 Hz
during a low force isometric contraction (inset shows the single unit EMG with greater temporal
resolution).
THE MANY FACES OF EMG: APPLICATIONS AND INTERPRETATION

Kinesiological EMG
 Muscle Activation Patterns: EMG can be used to determine whether a muscle is

simply active (on) or inactive (off). It has been utilized in this fashion by
biomechanists for decades to characterize the timing of muscle activation patterns
associated with various activities of daily living (e.g., walking), exercises (e.g.,
cycling), and ergonomic tasks (e.g., lifting). For applications of this nature surface
EMG is most commonly used, with the EMG onset typically being defined as the
point when the EMG amplitude exceeds a given threshold level that is based on a set
amount above baseline levels (e.g., 2-3 x greater than baseline noise for a specified
amount of time) (Thomas et al., 2007). It is also relatively common for the onset
latency of certain muscles to be calculated relative to those of other muscles using
similar criteria. In addition to using EMG to determine whether a muscle is active or
not, it can also be used in a variety of other more sophisticated ways to evaluate
muscle activation patterns during certain tasks. For example, the relative degree
(magnitude) of activation of a given muscle relative to a maximum contraction or
other agonistic and antagonistic muscles can be determined (Clark et al., 2005; Clark
et al., 2003a; Clark et al., 2003b; Hunter et al., 2004; Hunter and Enoka, 2001), as
well as whether EMG bursting is present (transient changes in the interference EMG
signal during a constant force voluntary task) (Clark et al., 2005; Kouzaki et al.,
2002; Kouzaki et al., 2003).
 Single Motor Unit Behavior: Using concentric needle and/or fine-wire EMG it is
possible to observe and record activity of single motor units especially during lower
force (< 20% of maximal strength) contractions (Figure 5B). Fine-wire EMG is most
commonly performed by threading two sterilized, insulated fine-wires with hooked

ends that have exposed bare wire (0.05 mm) through a needle (25-27 gauge) that is
inserted into a muscle. Following insertion, the needle is removed and the wires are
left embedded in the muscle tissue with the small exposed ends serving as the
recording electrodes until they are removed. Accordingly, the detection area of the
fine wires is small, and studying the behavior pattern of individual motor units under
various environmental and pathological conditions is possible. The most commonly
quantified outcomes of single motor unit activity involves calculating the average
time interval between a single motor unit firing to determine its discharge rate (de
Luca et al., 1996; Farina et al., 2009), calculating the discharge rate variability
(Kornatz et al., 2005; Moritz et al., 2005; Tracy et al., 2005), and defining motor unit
recruitment thresholds (Farina et al., 2009; Riley et al., 2008). In recent years
investigations on single motor unit behavior have provided valuable insight on motor
unit adaptations associated with motor control and aging (Enoka et al., 2003).
 Muscle Fiber Membrane Properties: The compound muscle fiber action potential
(CMAP) (Figure 5A), recorded with either surface or intramuscular electrodes, is of
interest in studying changes in muscle membrane properties. In general, CMAP‘s are
elicited by delivering a single, supramaximal electrical stimulation pulse (0.5-1 msec
in duration) to a peripheral nerve and recording the EMG response (Clark et al.,
2007; Clark et al., 2006a; Russ and Kent-Braun, 2003). The CMAP represents the
summated electrical activity (arising from motor unit action potentials) resultant of
the synchronous depolarization of the muscle fiber innervated by the depolarized
nerve. Thus, the CMAP waveform is determined by the effectiveness of temporal and
spatial summation which is affected by various factors. Changes in the CMAP‘s
amplitude is influenced by various factors such as the muscle fiber conduction
velocity, number of activated motor units, intracellular, while the CMAP duration is
mostly notably affected by temporal summation r action potential characteristics
(Keenan et al., 2006b).
 Spinal Reflexes and Motoneuron Excitability: The EMG activity associated with
electrical stimulation of peripheral nerves has also been utilized to investigate in vivo
excitability of the spinal reflexes. The most common approach to investigating spinal
reflex function has involved measuring the H-reflex (Zehr, 2002). The H-reflex is
evoked by electrically stimulating the peripheral nerve which elicits action potentials
in the sensory Ia afferents that propagate to the spinal cord where they give rise to
excitatory postsynaptic potentials and activate α-motoneuron axons (Figure 6). As
such, the amplitude of the EMG reflex response provides a global measure of spinal
excitability as it can be modified by a number of factors such as presynaptic
inhibition, the amount of Ia neurotransmitter released, and the excitability of the α-
motoneurons. The H-reflex (along with the stretch reflex which is similar to the H-
reflex except for its induction thru rapid stretch of the muscle spindle fibers) have
been and will continue to be powerful investigative tools that are centrally dependent
upon EMG (Aagaard et al., 2002; Clark et al., 2006b; Shinohara et al., 2005).
Additionally, recent methodological developments indicate that magnetic and
electrical stimulation at the level of the cervicomedullary junction can evoke single
descending volleys that can be detected with EMG recordings (Taylor, 2006). These
cervicomedullary evoked potentials activate α-motorneuron axons primarily through
a monosynaptic connection, and have been suggested to more directly assess α-

motorneuron excitability in vivo without the confounds inherent in the H-reflex
(Martin et al., 2006a; Martin et al., 2006b; Taylor, 2006; Taylor and Gandevia, 2004;
Zijdewind et al., 2006).
Figure 6. Neuropathway of the Hoffmann (H) reflex response and that of cervicomedullary junction
stimulation. The H-reflex is evoked by electrically stimulating the peripheral nerve which elicits action
potentials in the sensory Ia afferents that propagate to the spinal cord where they give rise to excitatory
postsynaptic potentials and activate alpha-motoneuron axons. As such, this reflex response provides a
global measure of spinal excitability as it can be modified by a number of factors such as presynaptic
inhibition, the amount of Ia neurotransmitter released, and the excitability of the α-motorneurons.
Conversely, magnetic stimulation at the level of the cervicomedullary junction evokes single
descending volleys which activate alpha-motorneuron axons primarily through a monosynaptic
connection, and can be used to more directly assess α-motorneuron excitability in vivo. Figure modified
from Aagaard et al., J Appl Physiol, 92: 2309-2318, 2002. Used with permission.
 Motor Cortical Function and Excitability: Transcranial magnetic stimulation (TMS)

and transcranial electrical stimulation (TES) can be used to activate the human motor
cortex and assess the integrity of the central motor pathways (Kobayashi and
Pascual-Leone, 2003; Reis et al., 2008; Ridding and Rothwell, 2007). While both
methods are non-invasive, TMS is also relatively painless and as such it has rapidly
increased in popularity and use. TMS is based on the principle of electromagnetic
induction, where a pulse of current passing through a coil placed over a person's head
creates a rapidly changing magnetic pulse that penetrates the skull and induces a
secondary ionic current in the brain (Figure 7A). When the stimulus intensity is of
sufficient strength to depolarize a sufficient number of descending neurons a motor
action will occur in the stimulated muscles and an evoked EMG response can be
detected (commonly referred to as motor evoked potentials [MEP‘s]). While TMS
was introduced in humans more than 20-years ago (Barker et al., 1985), its use to
study human cortical physiology has dramatically increased since the turn of the
century.
Figure 7. A: Transcranial magnetic stimulation (TMS) induces electrical currents in excitable tissue by
electromagnetic induction (here being demonstrated in the brain). B: TMS to the cortex during a muscle
contraction produces a motor evoked potential (MEP) followed by electrical quiescence before activity
resumes that is indicative of corticospinal inhibition and commonly referred to as the silent period. C:
The paired-pulse TMS technique involves coupling a conditioning stimulus with a test stimulus at
different intervals. Specifically, the intensity of a conditioning pulse is set below motor threshold (the
intensity where an MEP is not elicited), and the test pulse is set to a suprathreshold level. At short
interstimulus intervals (e.g., 3-msec) the conditioning stimulus inhibits the MEP, whereas at longer
interstimulus intervals (e.g., 15-msec) it facilitates the MEP. Paired pulses with inter-stimulus intervals
(ISI‘s) between 1–5 milliseconds results in short-interval intracortical inhibition (SICI) and it provides
a means of studying the activity of GABAA inhibitory circuits within the primary motor cortex
(decreases the MEP amplitude in comparison to a control pulse), whereas ISI‘s between 10-25
milliseconds results in intracortical facilitation (ICF) and it allows for the study of intracortical
facilitation that is controlled by GABAA and NMDA receptors.
 From an EMG methodological perspective MEP‘s are generally recorded in the same
fashion as other EMG signals, and to the untrained eye appear similar to any evoked
potential (i.e., a compound EMG potential). However, because the EMG signal is
being elicited at the level of the brain the physiological underpinnings are more
complicated than those of other responses. For example, the CMAP characteristics
are primarily ascribed to those of the muscle cell membrane, whereas MEP
characteristics are influenced by the muscle cell membrane properties along with
cortical and spinal excitability. Because there are many applications and outcomes
possible with brain stimulation, a complete discussion is beyond the scope of this
chapter, and readers are referred to excellent reviews on this topic area for more
detailed information (Kobayashi and Pascual-Leone, 2003; Reis et al., 2008; Ridding
and Rothwell, 2007). However, we will briefly provide an overview description of
some of the more common applications of TMS that incorporates the evaluation of
and quantification of MEP‘s recorded using EMG.
Motor threshold refers to the lowest TMS intensity necessary to evoke MEP‘s in the
target muscle when single-pulse stimuli are applied to the motor cortex. Motor threshold is
generally defined as the lowest intensity required to elicit MEP‘s of a set threshold level (e.g.,
> 50 μV peak-to-peak amplitude in at least 50% of trials at rest). Changes in MT can reflect
changes at a variety of levels (i.e., the neural membrane, axonal electronic properties, the
structure and number of excitatory projections onto the primary motor cortex, or upregulation
of receptors in this region) and hence represents a global assessment of the membrane
excitability of pyramidal neurons (Maeda and Pascual-Leone, 2003; Ziemann, 2004). The
MEP amplitude is also a commonly quantified variable. When TMS is applied to the motor
cortex at an intensity above MT high-frequency indirect waves (I waves) are elicited in the
corticospinal tract (Di Lazzaro et al., 2004). These waves are modifiable by many
mechanisms including neurotransmitters (i.e., glutatmate, GABA), modulators of
neurotransmission (i.e., acetylcholine, norepinephrine, and dopamine), and interneurones
contacted by corticospinal tract cells which all function to influence the amplitude of the MEP
(Figure 7B and 7C) (Maeda and Pascual-Leone, 2003; Ziemann, 2004). The silent period is a
duration of electrical quiescence following an MEP when TMS is performed during a
voluntary muscle contraction (Figure 7B) (Damron et al., 2008). There are several
mechanisms thought to contribute to the silent period, with spinal inhibitory mechanisms
thought to be active in the early part and the latter part being specifically cortical in its origin
and most likely mediated by GABAergic and dopaminergic cortical inhibitory mechanisms
(Brasil-Neto et al., 1992; Fuhr et al., 1991; Inghilleri et al., 1993; Ziemann et al., 1993).
In addition to the aforementioned parameters evoked with a single TMS pulse applied to
the motor cortex, coupling two pulses (paired-pulse TMS) can also be used to modify evoked
MEP‘s. Here, a conditioning stimulus is delivered in combination with a test stimulus at
different intervals. Specifically, the intensity of a conditioning pulse can be set below motor
threshold, and the test pulse is set to a suprathreshold level to study intracortical inhibitory
and facilitatory processes. Paired pulses with inter-stimulus intervals (ISI‘s) between 1–5
milliseconds results in short-interval intracortical inhibition and it provides a means of
studying the activity of GABAA inhibitory circuits within the primary motor cortex (decreases
the MEP amplitude in comparison to a control pulse), whereas ISI‘s between 10-25
milliseconds results in intracortical facilitation and it allows for the study of intracortical
facilitation that is controlled by GABAA and NMDA receptors (Figure 7C) (Kobayashi and
Pascual-Leone, 2003; Reis et al., 2008). Further, longer interstimulus intervals (50-200
milliseconds) with suprathreshold conditioning and test stimuli also results in inhibition (long
interval intracortical inhibition) (Kobayashi and Pascual-Leone, 2003; Reis et al., 2008).
While the underlying mechanisms of long interval intracortical inhibition are not fully
understood, it is thought that it is mediated within the primary motor cortex rather than
subcortical structures (Nakamura et al., 1997). Pharmacological studies indicate that long
interval intracortical inhibition is mediated by GABAB receptors (McDonnell et al., 2006;
Werhahn et al., 1999), and is likely to be mechanistically linked to the silent period.
 Mapping cortical function and reorganization: Since 1991, TMS evoked motor
responses have also been used to map brain functions in a direct stimulus/evoked
response manner previously only possible during invasive surgery when the surface
of the brain was exposed (Cohen et al., 1991; Penfield and Boldrey, 1937; Sohn and
Hallett, 2004). During cortical mapping, a grid is placed on the scalp (e.g., a swim
cap with a grid pattern) and the MEP amplitudes evoked at numerous sites are
determined and the values are plotted to create a 3-dimensional representation
between spatial location (x and y axis‘) and MEP amplitude (z-axis) (Figure 8)
(Thickbroom and Mastagliam, 2002). These cortical maps provide three pieces of
information: the total area on the scalp from which MEP‘s for the target muscle were
recorded, the “hot spot” for a muscle (the location where the largest MEP is
observed), and the amplitude weighted center of gravity (COG) (Wolf et al., 2005).
The COG corresponds to the center of the TMS map or the scalp location/topography
where the most neurons can be activated for a muscle or a movement which may or
may not be equivalent to the hot spot (Butler and Wolf, 2007; Thickbroom and
Mastagliam, 2002). Shifts in the location of COG (medial lateral or anterior posterior
directions) are commonly suggested to demonstrate cortical reorganization or
plasticity in response to injury, spontaneous recovery, or due to rehabilitation
intervention (Butler and Wolf, 2007; Curra et al., 2002).
Figure 8. 3-D contour map of cortical representation of skeletal muscle plotting the motor evoked
potential amplitude relative to the spatial location of the TMS coil. The x–y grid represents the surface
of the contralateral scalp, marked into 0.5 cm squares. The z-axis represents the peak to peak EMG
amplitude (expressed relative to Mmax) evoked at each point on the scalp for 10 stimuli.
These cortical maps, while insightful, need to be interpreted cautiously. Although the
stimulation protocol is similar to the principles used by Penfield, it is important to recognize
that the maps created using this technique do not compare in precision to maps created using
intracortical microstimulation (Butler and Wolf, 2007; Thickbroom and Mastagliam, 2002).
Animal studies have demonstrated that individual corticospinal neurons innervate several
motor neuron pools and thus different muscles and corticospinal neurons that innervate a
particular muscle are distributed among other corticospinal neurons projecting to different
muscle combinations (Nudo, 2006; Rossini and Dal Forno, 2004). This mosaic somatotopy of
the cortex and the overlapping spinal cord projections in combination with the lack of
stimulus precision with TMS means that multiple muscles will respond to a single TMS pulse
delivered at one point on the scalp matrix (Thickbroom and Mastagliam, 2002). The maps
usefulness can be further confounded by electrode placement that permits cross talk, or
signals evoked at the same time from other muscles, to interfere with the specificity and
quality of the recorded MEP (Wolf et al., 2005).
 Muscle Fatigue: Skeletal muscle fatigue has fascinated physiologists for more than a
century (Mosso, 1904). While muscle fatigue is associated with a variety of changes
in physiological properties- many of these properties are electrophysiological in
nature (or at a minimum manifest themselves with acute adjustments in
electrophysiological properties) (DeLuca, 1985; Hagg, 1992; Merletti et al., 2001).
For example, during a sustained, submaximal fatigue task a gradual increase in the
amplitude of the interference EMG signal is observed that is primarily due to an
increase in the motor units recruited as fatigue progresses (Figure 9A) (Basmaijian
and DeLuca, 1985). Additionally, during a sustained contraction the depolarization
and propagation of muscle fiber action potentials are modified. These modifications
produce time-dependent changes in the surface EMG signal, which result in a shift of
the power density spectrum to the lower frequencies (spectral compression) (Figure
9B) (Hagg, 1992). Spectral compression during a fatiguing submaximal contraction
has been attributed to a number of underlying physiological factors. One of the most
popular hypotheses states that the decrease in muscle fiber conduction velocity seen
with fatigue influences the power density spectrum, resulting in spectral compression
(Eberstein and Beattie, 1985; Lindstrom et al., 1977; Lindstrom et al., 1970; Merletti
et al., 1984). This is most likely due to an accumulation of metabolites (i.e., hydrogen
ion and extracellular potassium) (Bigland-Ritchie et al., 1981; Juel, 1988; Tesch et
al., 1983), reducing intracellular pH (Brody et al., 1991) and, thus, decreasing
sarcolemma excitability. However, this explanation appears to be incomplete, as a
disassociation between median frequency and conduction velocity is observed during
ischemia and different types of muscular contractions (Masuda et al., 1999; Zwarts et
al., 1987). In addition to the changes in the interference EMG signal that are of
interest for the study of muscle fatigue, many of the evoked EMG responses
described previously are altered. For example, there is evidence that fatigue
decreases the CMAP (Fuglevand et al., 1993) and prolongs the silent period
(Gandevia and Taylor, 2006; Taylor et al., 2006). As such, EMG is a valuable tool to
assess electrophysiological manifestations of fatigue.
Figure 9. Changes in surface EMG characteristics during a sustained, submaximal contraction. A: Note
the gradual increase in the amplitude of the interference EMG signal (top trace) as the force is
continually maintained to task failure at 25% of maximum strength (bottom trace of panel A). B: The
power density spectrum shifts towards the lower frequencies in association with fatigue. At the start of
the contraction the median frequency was 88 Hz (gray trace) and by the end of the contraction this was
reduced to 38 Hz (black trace).
Clinical Applications
EMG began to be widely used for clinical applications in the 1950‘s, and today it is still
commonly utilized in many fields of medicine including neurology, neurosurgery, physical
therapy, and physical medicine and rehabilitation. In the clinical context, patients with
symptoms suggestive of neuromuscular pathology are frequently referred to a neurologist or a
clinical neurophysiologist/electromyographer who conducts a variety of tests using EMG.
The term ―EMG‖ is commonly used by physicians and healthcare professions to refer to these
specific tests; however, as stated earlier- the term EMG should be used to simply refer to the
recording of electrical activity from the muscle fiber membrane. These tests frequently
involve both surface and intramuscular recordings from resting muscles, during voluntary
contraction as well as when electrically evoked and the information (EMG characteristics and
responses) can provide insight into the pathologic mechanism, severity, and anatomical
location of certain diseases (Daube and Rubin, 2009). Additionally, for the patient with a
known neurologic diagnosis, EMG is frequently used to monitor progression, to predict long
term potential for functional recovery, to evaluate response to therapy and to elucidate
underlying mechanisms of recovery (Gilchrist and Sachs, 2004; Patikas et al., 2005; Thomas
and Noga, 2003; van Kuijk et al., 2005). Depending upon the techniques used and the
understanding of normal neuromuscular physiology, the source of change in the EMG signal
can be localized to the muscle, neuromuscular junction, peripheral nerve, spinal nerve root,
spinal cord, or supraspinal structure based upon signature alterations to the EMG signal
(Daube and Rubin, 2009; Sella, 2003). Accordingly, it is imperative that clinicians also be
familiar with the technical and non-neuromuscular physiological influences on the EMG
signal in order to be certain that the differences identified in the recorded EMG signal are
truly due to an underlying pathological process and not due to an artifact of the recording and
analysis procedures (Daube and Rubin, 2009; Swash, 2002). Below we will briefly describe
some of the most common electrodiagnostic EMG tests as well as several other EMG
applications that are clinically relevant.
 Nerve Conduction Studies: One of the most common clinical uses of EMG is to study
the conduction properties of the peripheral neuromuscular system (Katirji, 2002).
These are frequently conducted to diagnose conduction disorders such as
radiculopathies and neuropathies. Conduction of sensory, mixed and motor nerves
are all commonly assessed, and in general utilize surface EMG recordings and
peripheral nerve stimulation where the latency of evoked EMG signals (e.g., M-
waves or reflex waves) are evaluated or stimulation is applied at two peripheral nerve
locations and the temporal separation of the two actions potentials relative to the
distance between the two stimulation sites is determined. However, it should also be
noted that from a clinical perspective measurements of interest associated with nerve
conduction studies typically extend beyond simple velocities and latencies and EMG
signal amplitudes and durations are also of interest (Katirji, 2002).
 Needle EMG: While we feel the term ‗needle EMG‘ is inappropriate and too non-
descriptive to truly be associated with a particular diagnostic test- it is nonetheless
commonly used to describe a relatively subjective examination where EMG is
recorded and evaluated for spontaneous and insertional activity, as well as the
general appearance and characteristics of single motor unit action potentials during
low level voluntary contractions (Katirji, 2002). In normal, healthy muscle there is
little or no EMG activity under resting conditions; however, in certain disorders (e.g.,
ALS, polymyositis) spontaneous activity at rest and abnormal action potentials
during contraction may be observed (Daube and Rubin, 2009; Echternach, 1994;
Katirji, 2002). Additionally, this clinical test can be suggestive of denervation which
is commonly associated with increased EMG fibrillations, positive sharp waves, and
giant motor unit action potentials (Daube and Rubin, 2009). While the needle EMG
recordings can provide valuable clinical insight, it should be noted that one of the
limitations of these examinations is its dependence of the skill of the examiner as
many of the outcomes are qualitative in nature.
 Motor Unit Number Estimation: Quantifying the number of motor axons innervating
a muscle (or muscle group) is of clinical importance for diagnosing and monitoring
the progression of a number of neurological diseases (e.g., amyotrophic lateral

sclerosis, spinal muscle atrophy). However, counting individual motor units (as
would be done in cadaveric studies) is not possible in vivo, but through the utilization
of EMG recordings estimates can be attained. This EMG technique, referred to as
motor unit number estimation (MUNE) is calculated based on a simple ratio of the
CMAP divided by the average surface motor unit potential (Bromberg, 2007). The
‗surface motor unit potential‘ is quantified by stimulating the peripheral nerve at
different intensities and evaluating the response variability of different groups of
axons. There are a number of technical issues and limitations associated with the
methods of the MUNE technique, and for a further discussion of these the reader is
referred to Bromberg (Bromberg, 2007).
 Biofeedback: Another clinical use of EMG is to provide biofeedback, generally based
on the degree of surface signal amplitude, as an indicator muscle activity. The use of
EMG biofeedback has been used for a wide variety of different applications mainly
aimed at ‗teaching‘ people how to exert voluntary control of their muscles (both to
increase and decrease the muscle activation depending on the rationale for the EMG
biofeedback). For example, it has been used in urinary incontinence programs to
provide feedback to patients regarding the activation of their pelvic floor muscles
(Dannecker et al., 2005). Additionally, it has been used in ergonomic applications to
provide feedback to patients to reduce the overall EMG activity of their trapezius
muscles during occupational computer work (Holtermann et al., 2008).
 Changes in muscle activation patterns associated with movement dysfunction: As
described above in the section Muscle Activation Patterns, clinicians use both surface
and needle electrodes to detect integrated EMG signals to detect changes in
activation patterns of multiple muscles during functional activities such as gait and
shoulder elevation (Jaggi et al., 2009; Winter, 1987). The relative onset/offset timing,
duration, and sequences of muscle EMG activity can be compared to data generated
by biomechanists to identify the muscular sources for the observed kinematic
changes during motion which could be caused by musculoskeletal injury or pain
(Cram and Kasman, 1998; De Luca, 1997; Thomas et al., 2007).
 Other Clinical Applications: There are numerous other clinically relevant
applications using EMG that have yet to gain widespread use. For example, clinical
applications of transcranial magnetic stimulation have begun to gain attention in
recent years for a variety of possibilities, including its use to monitor injury to central
motor pathways during surgery by examining changes in MEP‘s (Calancie and
Molano, 2008), as well diagnosing diseases such as ALS, myelopathies, and multiple
sclerosis through the evaluation of the evoked EMG responses (e.g., silent period,
central motor conduction time) (Chen et al., 2008). TMS methods are frequently used
to predict the potential for functional recovery early after hemispheric stroke,
especially in the upper extremity and hand (Chen and Winstein, 2009; Curra et al.,
2002). Even when the patient presents with significant clinical deficits, the presence
and amplitude of an MEP in the acute phase has a high positive predictive value for
restoration of function whereas absence of or prolonged MEP latency and conduction
times predict poor return of hand motor ability (Butler and Wolf, 2007).
Additionally, EMG has been, and will continue to be evaluated as a potential marker
for other diseases and disorders as illustrated by the recent work in evaluating EMG
amplitude of the submentalis muscle in sleep behavior disorders (Feri et al., 2008).
Thus, it seems that the continual growth and expansion of clinical applications of
EMG will continue to be developed in the 21st century.
PERILS OF EMG AND CHAPTER SUMMARY

EMG has been used to expand our knowledge base regarding human movement and
neuromuscular disorders for many years. This chapter provides an overview of the
fundamentals for a basic understanding and appreciation of electromyography and its many
applications. However, we must caution that there are many pitfalls and perils associated with
these applications that must be understood for appropriate research design and data
interpretation (Basmaijian and DeLuca, 1985; Farina, 2006; Farina et al., 2004). For example,
as stated earlier, the absolute amplitude of EMG signals is of trivial consequence when
comparisons are being made between individuals, muscles and even over time (Lehman and
McGill, 1999). Even normalizing the EMG to a maximal voluntary contraction has
limitations due to the assumption that the maximal voluntary contraction is not associated
with any failure in central activation. One approach to circumventing this issue would be to
normalize the EMG data to an electrically stimulated maximal contraction, but this requires
that a peripheral nerve be accessible (as direct muscle stimulation will result in too much
artifact to allow for recording stable CMAP‘s) which is not feasible in many muscles. Other
perils of using EMG to evaluate muscle activation patterns are the difficulty in dealing with
movement/wire artifact, issues related to cross-talk (signals detected from nearby muscles), as
well as issues surrounding interpreting EMG during dynamic tasks where the detection area
of underlying muscles changes in relation to surface electrodes located on the skin as the
muscle shorten and lengthens during movement (Farina, 2006). Further, day-to-day variation
in electrode placement, temperature and state (state-dependency) can confound longitudinal
studies. However, despite its limitations, when used appropriately EMG can provide a
window into the state of the human neuromuscular system.
ABOUT THE AUTHORS

Petra Williams is an Assistant Professor in the School of Physical Therapy at Ohio
University, a practicing physical therapist for 15-years, and a board-certified Neurologic
Clinical Specialist. Her clinical and research interest include the mechanisms, diagnosis and
treatment of movement dysfunction associated with neuromuscular injury. She has received a
B.A. in History and Literature of Religions from Northwestern University, and a certificate of
Physical Therapy and a M.S. in Advanced Professional Neurologic Practice from The Ohio
State University, She is currently pursuing a Ph.D. in Comparative and Exercise Physiology
through the Department of Biological Sciences at Ohio University with a focus on
neuromuscular adaptation.
Brian Clark is the Director of the Institute for Neuromusculoskeletal Research at Ohio
University, where he is an Assistant Professor in the Department of Biomedical Sciences
housed in the College of Osteopathic Medicine. He also serves on the Graduate Faculty in the
Comparative and Exercise Physiology Program in the Department of Biological Sciences at
Ohio University. He has received a B.S. in Biology from Western Carolina and an M.S. and
Ph.D. in Exercise Physiology from Syracuse University. His research interests surround the
neuromuscular mechanisms that mediate acute adjustments (e.g., muscle fatigue) and chronic
adaptations (e.g., cast immobilization, exercise) in response to physical activity and under
pathological conditions. He has over 80 publications (40+ articles and 40+ abstracts) on
neuromuscular physiology and EMG. His research has been supported by grants from NASA
and the American College of Sports Medicine.
REFERENCES
Aagaard, P., E. B. Simonsen, J. L. Andersen, et al. Neural adaptation to resistance training:
changes in evoked V-wave and H-reflex responses J. Appl. Physiol. 92(6): 2309-18,
2002.
Barker, A. T., R. Jalinous and I. L. Freeston. Non-invasive magnetic stimulation of human
motor cortex Lancet 1(8437): 1106-7, 1985.
Basmaijian, J. V. and C. L. DeLuca. Muscle Alive- their functions revealed by
electromyography. Baltimore, Williams and Wilkins, 1985.
Bigland-Ritchie, B., E. F. Donovan and C. S. Roussos. Conduction velocity and EMG power
spectrum changes in fatigue of sustained maximal efforts J. Appl. Physiol. 51(5): 1300-5.,
1981.
Brasil-Neto, J. P., L. G. Cohen, M. Panizza, et al. Optimal focal transcranial magnetic
activation of the human motor cortex: effects of coil orientation, shape of the induced
current pulse, and stimulus intensity J. Clin. Neurophysiol. 9(1): 132-6, 1992.
Brody, L. R., M. T. Pollock, S. H. Roy, et al. pH-induced effects on median frequency and
conduction velocity of the myoelectric signal J. Appl. Physiol. 71(5): 1878-85., 1991.
Bromberg, M. B. Updating motor unit number estimation (MUNE) Clin. Neurophysiol.
118(1): 1-8, 2007.
Butler, A. J. and S. L. Wolf. Putting the brain on the map: use of transcranial magnetic
stimulation to assess and induce cortical plasticity of upper-extremity movement Phys.
Ther. 87(6): 719-36, 2007.
Calancie, B. and M. R. Molano. Alarm criteria for motor-evoked potentials: what's wrong
with the "presence-or-absence" approach? Spine 33(4): 406-414, 2008.
Cavanagh, P. R. and P. V. Komi. Electromechanical delay in human skeletal muscle under
concentric and eccentric contractions Eur J. Appl. Physiol. Occup Physiol 42(3): 159-63,
1979.
Chen, R., D. Cros, A. Curra, et al. The clinical diagnostic utility of transcranial magnetic
stimulation: report of an IFCN committee Clin. Neurophysiol. 119(3): 504-532, 2008.
Chen, S. Y. and C. J. Winstein. A systematic review of voluntary arm recovery in hemiparetic
stroke: critical predictors for meaningful outcomes using the international classification
of functioning, disability, and health J. Neurol. Phys. Ther. 33(1): 2-13, 2009.
Clark, B. C., S. R. Collier, T. M. Manini, et al. Sex differences in muscle fatigability and
activation patterns of the human quadriceps femoris Eur. J. Appl. Physiol. 94(1-2): 196-
206, 2005.
Clark, B. C., S. B. Cook and L. L. Ploutz-Snyder. Reliability of techniques to assess human
neuromuscular function in vivo J. Electromyogr. Kinesiol. 17(1): 90-101, 2007.
Clark, B. C., B. Fernhall and L. L. Ploutz-Snyder. Adaptations in human neuromuscular
function following prolonged unweighting: I. Skeletal muscle contractile properties and
applied ischemia efficacy J. Appl. Physiol. 101(1): 256-63, 2006a.
Clark, B. C., T. M. Manini, S. J. Bolanowski, et al. Adaptations in human neuromuscular
function following prolonged unweighting: II. Neurological properties and motor
imagery efficacy J. Appl. Physiol. 101(1): 264-72, 2006b.
Clark, B. C., T. M. Manini and L. L. Ploutz-Snyder. Derecruitment of the lumbar musculature
with fatiguing trunk extension exercise Spine 28(3): 282-7, 2003a.
Clark, B. C., T. M. Manini, D. J. The, et al. Gender differences in skeletal muscle fatigability
are related to contraction type and EMG spectral compression J. Appl. Physiol. 94(6):
2263-72, 2003b.
Cohen, L. G., S. Bandinelli, T. W. Findley, et al. Motor reorganization after upper limb
amputation in man. A study with focal magnetic stimulation Brain 114 ( Pt 1B): 615-27,
1991.
Cram, J. R. and G. S. Kasman. Introduction to Surface Electromyography. Gaithersburg, MD,
Aspen Publishers, 1998.
Curra, A., N. Modugno, M. Inghilleri, et al. Transcranial magnetic stimulation techniques in
clinical investigation Neurology 59(12): 1851-9, 2002.
Damron, L. A., R. L. Hoffman, D. J. Dearth, et al. Quantification of the corticospinal silent
period evoked via transcranial magnetic brain stimulation J. Neurosci. Methods 173: 121-
128, 2008.
Dannecker, C., V. Wolf, R. Raab, et al. EMG-biofeedback assisted pelvic floor muscle
training is an effective therapy of stress urinary or mixed incontinence: a 7 year
experience with 390 patients Arch. Gynecol. Obstet. 273: 93-97, 2005.
Daube, J. R. and D. I. Rubin. Needle electromyography Muscle Nerve 39(2): 244-70, 2009.
De Luca, C. J. The use of surface electromyography in biomechanics Journal of Applied
Biomechanics 13: 135-163, 1997.
De Luca, C. J., P. J. Foley and Z. Erim. Motor unit control properties in constant-force
isometric contractions J. Neurophysiol. 76(3): 1503-16, 1996.
DeLuca, C. J. Myoelectric manifestations of localized muscular fatigue in humans Crit. Rev.
Biomed .Eng. (11): 251-279, 1985.
Di Lazzaro, V., A. Oliviero, F. Pilato, et al. The physiological basis of transcranial motor
cortex stimulation in conscious humans Clin. Neurophysiol. 115(2): 255-66, 2004.
Dumitru, D. Physiological basis of potentials recorded in electromyography Muscle and
Nerve 23: 1667-1685, 2000.
Eberstein, A. and B. Beattie. Simultaneous measurement of muscle conduction velocity and
EMG power spectrum changes during fatigue Muscle Nerve 8(9): 768-73., 1985.
Echternach, J. L. Introduction to Electromyography and Nerve Conduction Testing.
Thorofare, NJ, SLACK Incorporated, 1994.
Enoka, R. M. Neuromechanics of human movement. Champaign, Human Kinetics, 2002.
Enoka, R. M., E. A. Christou, S. K. Hunter, et al. Mechanisms that contribute to differences

in motor performance between young and old adults J. Electromyogr. Kinesiol. 13: 1-12,
2003.
Farina, D. Interpretation of the surface electromyogram in dynamic contractions Exerc. Sport
Sci. Rev. 34(3): 121-7, 2006.
Farina, D., A. Holobar, M. Gazzoni, et al. Adjustments differ among low-threshold motor
units during intermittent, isometric contractions J. Neurophysiol. 101(1): 350-9, 2009.
Farina, D., R. Merletti and R. M. Enoka. The extraction of neural strategies from the surface
EMG J. Appl. Physiol. 96(4): 1486-95, 2004.
Feri, R., C. Franceshini, M. Zucconi, et al. Searching for a marker of REM sleep behavior
disorder: submentalis muscle EMG amplitude analysis during sleep in patients with
narcolepsy/cataplexy Sleep 31(10): 1409-1417, 2008.
Fuglevand, A. J., K. M. Zackowski, K. A. Huey, et al. Impairment of neuromuscular
propagation during human fatiguing contractions at submaximal forces J. Physiol. 460:
549-572, 1993.
Fuhr, P., R. Agostino and M. Hallett. Spinal motor neuron excitability during the silent period
after cortical stimulation Electroencephalogr Clin. Neurophysiol. 81(4): 257-62, 1991.
Galvani, L. De viribus electricitatis in motu musculari: Commentarius. Bologna, Tip. Istituto
delle Scienze, 1791.
Gandevia, S. C. and J. L. Taylor. Supraspinal fatigue: the effects of caffeine on human muscle
performance J. Appl. Physiol. 100(6): 1749-50, 2006.
Gilchrist, J. M. and G. M. Sachs. Electrodiagnostic studies in the management and prognosis
of neuromuscular disorders Muscle Nerve 29(2): 165-90, 2004.
Hagg, G. M. Interpretation of EMG spectral alterations and alteration indexes at sustained
contraction J. Appl. Physiol. 73(4): 1211-7, 1992.
Holtermann, K., K. Sogaard, H. Christensen, et al. The influence of biofeedback training on
trapezius activity and rest during occupational computer work: a randomized controll trial
Eur. J. Appl. Physiol. 104: 983-989, 2008.
Hunter, S. K., A. Critchlow, I. Shin, et al. Fatigability of the elbow flexor muscles for a
sustained submaximal contraction is similiar in men and women matched for strength J.
Appl. Physiol. 96: 195-202, 2004.
Hunter, S. K. and R. M. Enoka. Sex differences in the fatigability of arm muscles depends on
absolute force during isometric contractions J. Appl. Physiol. 91(6): 2686-94., 2001.
Inghilleri, M., A. Berardelli, G. Cruccu, et al. Silent period evoked by transcranial stimulation
of the human cortex and cervicomedullary junction J. Physiol. 466: 521-34, 1993.
Jaggi, A., A. A. Malone, J. Cowan, et al. Prospective blinded comparison of surface versus
wire electromyographic analysis of muscle recruitment in shoulder instability Physiother
Res. Int. 14(1): 17-29, 2009.
Juel, C. Muscle action potential propagation velocity changes during activity Muscle Nerve
11(7): 714-9., 1988.
Kandel, E. R., J. H. Schwartz and T. M. Jessel. Principles of Neural Science. New York, Mc-
Graw Hill, 2000.
Katirji, B. The clinical electromyography examination. An overview Neurol. Clin. 20(2): 291-
303, v, 2002.
Keenan, K. G., D. Farina, K. S. Maluf, et al. Influence of amplitude cancellation on the
simulated surface electromyogram J. Appl. Physiol. 98(1): 120-31, 2005.
Keenan, K. G., D. Farina, R. Merletti, et al. Amplitude cancellation reduces the size of motor
unit potentials averaged from the surface EMG J. Appl. Physiol. 100(6): 1928-37, 2006a.
Keenan, K. G., D. Farina, R. Merletti, et al. Influence of motor unit properties on the size of
the simulated evoked surface EMG potential Exp. Brain Res. 169(1): 37-49, 2006b.
Kobayashi, M. and A. Pascual-Leone. Transcranial magnetic stimulation in neurology Lancet
Neurol 2(3): 145-56, 2003.
Kornatz, K. W., E. A. Christou and R. M. Enoka. Practice reduces motor unit discharge
variability in a hand muscle and improves manual dexterity in old adults J. Appl. Physiol.
98(6): 2072-80, 2005.
Kouzaki, M., M. Shinohara, K. Masani, et al. Alternate muscle activity observed between
knee extensor synergists during low-level sustained contractions J. Appl. Physiol. 93:
675-684, 2002.
Kouzaki, M., M. Shinohara, K. Masani, et al. Local blood circulation among knee extensor
synergists in relation to alternate muscle activity during low-level sustained contraction J.
Appl. Physiol. 95: 49-56, 2003.
Lehman, G. J. and S. M. McGill. The importance of normalization in the interpretation of
surface electromyography: a proof of principle J. Manipulative Physiol. Ther. 22(7): 444-
6, 1999.
Lindstrom, L., R. Kadefors and I. Petersen. An electromyographic index for localized muscle
fatigue J. Appl. Physiol. 43(341-353), 1977.
Lindstrom, L., R. Magnusson and I. Petersen. Muscular fatigue and action potential
conduction velocity changes studied with frequency analysis of EMG signals
Electromyography 10(4): 341-56., 1970.
MacIntosh, B. R., P. F. Gardiner and A. J. McComas. Skeletal Muscle: Form and Function.
Champaign, IL, Human Kinetics, 2006.
Maeda, F. and A. Pascual-Leone. Transcranial magnetic stimulation: studying motor
neurophysiology of psychiatric disorders Psychopharmacology (Berl) 168(4): 359-76,
2003.
Marey, E. J. Théorie du myographe qui inscrit le raccourcissement du muscle, 1876.
Martin, P. G., S. C. Gandevia and J. L. Taylor. Output of human motoneuron pools to
corticospinal inputs during voluntary contractions J. Neurophysiol. 95(6): 3512-8, 2006a.
Martin, P. G., J. L. Smith, J. E. Butler, et al. Fatigue-sensitive afferents inhibit extensor but
not flexor motoneurons in humans J. Neurosci. 26(18): 4796-802, 2006b.
Masuda, K., T. Masuda, T. Sadoyama, et al. Changes in surface EMG parameters during
static and dynamic fatiguing contractions J. Electromyogr. Kinesiol. 9(1): 39-46., 1999.
Mathews, G. G. Cellular Physiology of Nerve and Muscle. Malden, MA, Blackwell
Publishing Company, 2003.
McDonnell, M. N., Y. Orekhov and U. Ziemann. The role of GABA(B) receptors in
intracortical inhibition in the human motor cortex Exp. Brain Res. 173(1): 86-93, 2006.
Merletti, R., A. Rainoldi and D. Farina. Surface electromyography for noninvasive
characterization of muscle Exerc. Sport Sci. Rev. 29(1): 20-5, 2001.
Merletti, R., M. A. Sabbahi and C. J. De Luca. Median frequency of the myoelectric signal.
Effects of muscle ischemia and cooling Eur. J. Appl. Physiol. Occup Physiol 52(3): 258-
65, 1984.
Moritz, C. T., B. K. Barry, M. A. Pascoe, et al. Discharge rate variability influences the
variation in force fluctuations across the working range of a hand muscle J.
Neurophysiol. 93(5): 2449-59, 2005.
Mosso, A. Fatigue. London, Swan Sonnenschein, 1904.
Nakamura, H., H. Kitagawa, Y. Kawaguchi, et al. Intracortical facilitation and inhibition after
transcranial magnetic stimulation in conscious humans J. Physiol. 498 ( Pt 3): 817-23,
1997.
Norman, R. W. and P. V. Komi. Electromechanical delay in skeletal muscle under normal
movement conditions Acta. Physiol. Scand. 106(3): 241-8, 1979.
Nudo, R. J. Plasticity NeuroRx 3(10): 420-427, 2006.
Patikas, D., S. Wolf and L. Doderlein. Electromyographic evaluation of the sound and
involved side during gait of spastic hemiplegic children with cerebral palsy Eur. J.
Neurol. 12(9): 691-9, 2005.
Penfield, W. and E. Boldrey. Somatic motor and sensory representation in cerebral cortex of
man as studied by electrical stimulation Brain 60: 389-443, 1937.
Petersen, N. T., H. S. Pyndt and J. B. Nielsen. Investigating human motor control by
transcranial magnetic stimulation Exp. Brain Res. 152(1): 1-16, 2003.
Reis, J., O. B. Swayne, Y. Vandermeeren, et al. Contribution of transcranial magnetic
stimulation to the understanding of cortical mechanisms involved in motor control J.
Physiol. 586(2): 325-51, 2008.
Ridding, M. C. and J. C. Rothwell. Is there a future for therapeutic use of transcranial
magnetic stimulation? Nat. Rev. Neurosci. 8(7): 559-67, 2007.
Riley, Z. A., A. H. Maerz, J. C. Litsey, et al. Motor unit recruitment in human biceps brachii
during sustained voluntary contractions J. Physiol. 586(8): 2183-93, 2008.
Rossini, P. M. and G. Dal Forno. Integrated technology for evaluation of brain function and
neural plasticity Phys. Med. Rehabil. Clin. N. Am. 15(1): 263-306, 2004.
Russ, D. W. and J. A. Kent-Braun. Sex differences in human skeletal muscle fatigue are
eliminated under ischemic conditions J. Appl. Physiol. 94: 2414-2422, 2003.
Sella, G. E. Neuropathology considerations: clinical and SEMG/biofeedback applications
Appl Psychophysiol Biofeedback 28(2): 93-105, 2003.
Shinohara, M., C. T. Moritz, M. A. Pascoe, et al. Prolonged muscle vibration increases stretch
reflex amplitude, motor unit discharge rate, and force fluctuations in a hand muscle J.
Appl. Physiol. 99(5): 1835-42, 2005.
Sohn, Y. H. and M. Hallett. Motor evoked potentials Phys. Med. Rehabil. Clin. N. Am. 15(1):
117-31, vii, 2004.
Swash, M. What does the neurologist expect from clinical neurophysiology? Muscle Nerve
Suppl 11: S134-8, 2002.
Taylor, J. L. Stimulation at the cervicomedullary junction in human subjects J. Electromyogr.
Kinesiol. 16(3): 215-23, 2006.
Taylor, J. L. and S. C. Gandevia. Noninvasive stimulation of the human corticospinal tract J.
Appl. Physiol. 96(4): 1496-503, 2004.
Taylor, J. L., G. Todd and S. C. Gandevia. Evidence for a supraspinal contribution to human
muscle fatigue Clin. Exp. Pharmacol. Physiol. 33(4): 400-5, 2006.
Tesch, P. A., P. V. Komi, I. Jacobs, et al. Influence of lactate accumulation of EMG
frequency spectrum during repeated concentric contractions Acta. Physiol. Scand. 119(1):
61-7., 1983.
Thickbroom, G. W. and F. L. Mastagliam (2002). Mapping Studies. Handbook of

Transcranial Magnetic Stimulation. A. Pascual-Leone. London, Arnold Publishers.
Thomas, C. K. and B. R. Noga. Physiological methods to measure motor function in humans
and animals with spinal cord injury J. Rehabil. Res. Dev. 40(4 Suppl 1): 25-33, 2003.
Thomas, J. S., C. R. France, D. Sha, et al. The effect of chronic low back pain on trunk
muscle activations in target reaching movements with various loads Spine 32(26): E801-
8, 2007.
Tracy, B. L., K. S. Maluf, J. L. Stephenson, et al. Variability of motor unit discharge and
force fluctuations across a range of muscle forces in older adults Muscle Nerve 32(4):
533-40, 2005.
van Kuijk, A. A., J. W. Pasman, A. C. Geurts, et al. How salient is the silent period? The role
of the silent period in the prognosis of upper extremity motor recovery after severe stroke
J. Clin. Neurophysiol. 22(1): 10-24, 2005.
Werhahn, K. J., E. Kunesch, S. Noachtar, et al. Differential effects on motorcortical inhibition
induced by blockade of GABA uptake in humans J. Physiol. 517 ( Pt 2): 591-7, 1999.
Winter, D. A. The biomechanics of motor control and human gait. Ontario, Canada,
University of Waterloo Press, 1987.
Wolf, S. L., A. J. Butler, J. L. Alberts, et al. Contemporary linkages between EMG, kinetics
and stroke rehabilitation J. Electromyogr. Kinesiol. 15(3): 229-39, 2005.
Zehr, P. E. Considerations for use of the Hoffmann reflex in exercise studies Eur. J. Appl.
Physiol. 86(6): 455-68, 2002.
Ziemann, U. TMS and drugs Clin. Neurophysiol. 115(8): 1717-29, 2004.
Ziemann, U., J. Netz, A. Szelenyi, et al. Spinal and supraspinal mechanisms contribute to the
silent period in the contracting soleus muscle after transcranial magnetic stimulation of
human motor cortex Neurosci. Lett. 156(1-2): 167-71, 1993.
Zijdewind, I., J. E. Butler, S. C. Gandevia, et al. The origin of activity in the biceps brachii
muscle during voluntary contractions of the contralateral elbow flexor muscles Exp.
Brain Res. 175(3): 526-35, 2006.
Zwarts, M. J., T. W. Van Weerden and H. T. Haenen. Relationship between average muscle
fibre conduction velocity and EMG power spectra during isometric contraction, recovery
and applied ischemia Eur. J. Appl. Physiol. Occup. Physiol. 56(2): 212-6, 1987.
Chapter 7
BIOMECHANICS IN CHILDREN WITH CEREBRAL

PALSY
Jessie Chen1 and Dinah Reilly2

1
Department of Physiology and Neuroscience,
New York University School of Medicine, New York, NY, USA
2
Department of Occupational and Physical Therapy,
Boise School District, Boise, ID, USA
ABSTRACT
Children with cerebral palsy (CP) lack the higher-level motor skills present in age-
matched typically developing (TD) children. The development of postural control is
critical to the acquisition of increasingly complex motor skills as well as to the
production of coordinated motor behavior, such as locomotion. This chapter examines
recent developments in understanding the abnormal postural control in children with CP
and assessments of the effectiveness of rehabilitation techniques using biomechanics
measurements. We show that the delayed and impaired development of postural control
in children with CP is not only due to the immaturity of central nervous system but also
abnormal postural alignment and muscle force production.
Children with CP often have difficulty maintaining stability when facing unexpected
threat to balance. We present studies comparing reactive balance control in children with
spastic diplegic cerebral palsy (SDCP) and TD children using support surface
perturbation and show that a number of factors contribute to poor balance control in
children with SDCP. 1) There was a temporal disorganization of joint torque activation.
2) There was a spatial disorganization of the joint torque profiles. 3) Children with SDCP
also showed slower speed to reach peak torque value. In addition we show that when TD
children were asked to mimic crouched stance as that seen in children with SDCP, they
exhibited abnormal postural control as well, indicating that musculoskeletal constraints
are also contributors to the atypical postural muscle response patterns seen in children
with SDCP. These findings suggest that the neuromuscular response patterns of some
children with SDCP may be appropriate strategies for their musculoskeletal constraints
secondary to deficits in the neural system.
In this chapter we also discuss the culmination of our findings in relation to clinical
applications in the management of musculoskeletal impairments to improve postural
234 Jessie Chen and Dinah Reilly
control in children with SDCP, and the significance of using biomechanical measures to
show a direct relationship between the impairments of the musculoskeletal system and
reactive postural control as well as possible coping strategies used by children with
SDCP. We examine the current biomechanical research used to ascertain the
effectiveness of two therapeutic interventions purported to affect the musculoskeletal
system for the improvement of function in children with SDCP, ankle foot orthoses
(AFO) and strength training.
Finally, we examine the gaps in current clinical research when assessing the
effectiveness of interventions to reduce the musculoskeletal impairments constraining
static, reactive, and dynamic balance control in children with SDCP.
INTRODUCTION
Children with motor system impairments, such as those seen in cerebral palsy (CP), lack
the higher-level motor skills present in age-matched healthy children. It is generally
understood that cerebral palsy is a category of chronic non-progressive disorders of
movement and posture of early onset (Styer-Acevedo, 1994). Research in the last 20 years has
shown that delayed and impaired development of postural control in children with CP is
attributed to a number of factors, including an immature central nervous system, the
pathophysiology of the neuromuscular system, an impaired postural alignment, and muscle
weakness. Postural control is a fundamental motor function involved in nearly every motor
task. The development of postural stability is critical for the acquisition of increasingly
complex motor skills as well as for the production of coordinated motor behavior seen in the
developing child, i.e., walking, stair climbing, jumping, and skipping (Roncesvalles et al,
2001). There has been a rich body of literature studying the alterations of the neural system
for postural control in children with CP; however, few studies have examined the
biomechanics of the musculoskeletal system on the control of posture.
Most of our knowledge on the development of postural control comes from research
using the support surface perturbation paradigm, in which an unexpected perturbation is given
and the resulting behaviors are documented as well as the associated muscle activity profiles
captured by electromyography (EMG) (Nashner, 1976; Nashner, 1977; Horak and Nashner,
1986; Nashner and McCollum, 1985). Three basic coping strategies have been identified
based on the behavioral responses to varying magnitudes of perturbation: ankle, hip and
stepping strategies. It was suggested that the forms of automatic postural response elicited
following a translation of the support surface became more consistent and efficient with an
increase in chronological age (Sveistrup and Woollacott, 1996). In an adult, muscle responses
are activated in a distal-to-proximal sequence; i.e., ankle muscles activated prior to knee and
hip muscles.
Children with cerebral palsy exhibit exaggerated dynamic stretch reflex responses, a lack
of reflex suppression on repetitive stimulation, and abnormal radiation of activity to
functionally related muscles (Nashner et al, 1983; Burtner et al, 1998; Woollacott et al, 1998).
EMG recordings of muscle activity associated with postural responses to balance threat
showed an increase in antagonist recruitment and decreased trunk activation when compared
to typically developing children at the same level of walking experience. In addition to high
level of agonist and antagonist muscle co-activation, children with spastic cerebral palsy
exhibited temporal reversals among the muscles responding to a loss of balance.
Biomechanics in Children with Cerebral Palsy 235
Not only do children with CP show abnormal neural responses indicated by altered EMG
patterns, they also exhibit abnormality in biomechanical measures. Children with spastic
cerebral palsy frequently show restricted range of motion at the ankle, knee, and hip
secondary to abnormal muscle tone and disordered movement patterns. The habitual crouched
postural pattern during stance results in subsequent shortening of the hamstring muscles,
which in turn ensures the continued use of the habitual crouched posture (Shumway-Cook
and Woollacott, 2001).
From a biomechanical point of view, in the stance position, the human body can be
treated as an inverted pendulum. In its simplest version, the body is considered a single rigid
segment with all the external forces acting on the center of mass located near the waist level.
The task of postural control is to keep the body‘s center of mass (COM) within the base of
support (BOS), defined by the individual's foot length and stance width (Winter et al, 1990).
Throughout development, children face a series of morphological changes to their bodies,
including an increase in height and weight; but more importantly, a lowering of the center of
mass due to changes in relative mass of the different body parts. Therefore, the seemingly
simple task of postural and balance control becomes more complex for growing children as
the difficulty of adapting to these changes increases. Researchers using traditional theories of
developing postural control, such as reflex-hierarchical theory, find it difficult to uncover a
universal solution that governs successful postural control when dealing with changes in so
many dimensions.
In the past several decades motor control scientists have been faced with the immense
task of discovering how the brain calculates and coordinates a large number of kinetic
parameters (forces produced by individual muscle groups) in order to produce appropriate
kinematic outcomes (position, velocity and acceleration of individual joints, body segments,
and whole body movements) for postural control when facing an external balance threat.
Winter, a biomechanist and a researcher in motor control, recently put forth a revolutionary
concept using biomechanical models (Winter, 1995) in an effort to simplify this complex
issue by reducing the large number of independent parameters into an interaction between
two parameters. He proposed that the difference between the center of pressure (COP) (the
net pressure force on the body acting through a single point in the BOS) and the center of
gravity (COG) (the vertical projection of the center of mass on the ground) is proportional to
the acceleration of body sway, and thus is recognized by the balance control system as the
"error" signal. Therefore, the greater the COG-COP separation, the faster the acceleration of
the COG, and the greater the demand on the postural control system. In response to the
balance threat, synergistic muscle responses are triggered automatically in peripheral
effectors to keep the COG within BOS. When a corrective movement of the center of pressure
to the extremes of the base of support (the edges of the feet) is not adequate to reverse the
acceleration of the center of gravity, a step is necessary, or falling occurs.
For children to maintain an upright, quiet stance, the nervous system must detect the
deviation of COG away from the vicinity of COP, and then generate sufficient torques around
the ankle, knee, and hip joints to counterbalance the tipping force. In dynamic balance, i.e.,
when walking or climbing stairs, this signal is not only used to keep balance, but used to
provide forward momentum as well. The difference between the COG and COP, although a
concept borrowed from the field of biomechanics, is considered a sensitive indicator by motor
control researchers in speculating the capability of the nervous system to control balance.
It is our understanding that in order to examine the development of balance control,

multiple levels of analyses are required. Kinematic data record the motor outcomes. EMG
recordings provide insight into the response from the central nervous system, because there is
a one-to-one correlation between output from spinal motor neurons and muscle contractions.
These two analyses alone, however, are not sufficient to represent the internal and external
forces acting on the joints or the whole body to explain the behavioral outcome. As the
singular measure of motor responses, the surface EMG is plagued by at least three significant
limitations (Bothner, 1998). 1) It can only record from the more superficial muscle groups
involved in the task, therefore, an assumption that these are the only muscles involved in
recovery of balance is an oversimplification of the total muscle function. 2) No direct
correlation has been found between EMG activity and force produced during postural
responses, particularly in a reactive balance control task such as postural recovery from a
perturbation involving a number of joints. Linear relations have only been observed in single
joint movements under unique circumstances such as isometric or isokinetic contractions. 3)
The direct relationship between EMG signals and behavior under dynamic circumstances
(such as walking and running) remains inconclusive.
Kinetic analyses, on the other hand, have been proposed as a necessary and
comprehensive approach to uncover control variables and strategies (Winter, 1995). Kinetic
analysis estimates the overall muscle forces around a joint responsible for an observed
movement, yielding variables that describe the causes of a motion. Analysis of joint torque, in
particular, provides a description and a kinetic account not available through kinematics and
EMG analysis alone. In this chapter, we will show results from studies of postural control in
children with CP using joint torque analyses and the resultant control of posture by looking at
the relationship between COG (kinematic parameter) and COP (kinetic parameter).
PURPOSE OF THIS CHAPTER

Research on postural control in children with cerebral palsy has examined four types of
postural tasks: 1) steady state postural control, in which the child simply maintains quiet
stance balance for a specific period of time; 2) reactive postural control, in which the child
reacts to an unexpected threat to balance, like a slip or trip; 3) proactive postural control, in
which the child integrates postural responses into voluntary tasks, and 4) dynamic postural
control, in which the relationship between the COG and COP continuously changes
(locomotion, climbing stairs, stepping over obstacles). Based upon our investigation, it is
clear that previous research lacks information on the kinetic changes and musculoskeletal
constraints associated with the development of reactive balance control in children with
cerebral palsy.
Our group has attempted to fill the gaps by examining the differences in kinetic profiles
of the lower extremity joints during reactive postural control in children with different types
of CP (diplegic, hemiplegic and ataxic) and typically developing children (Chen and
Woollacott, 2007).
Our primary purpose of this section is to examine how torque productions around lower
extremity joints of children with spastic diplegic cerebral palsy (SDCP) differ from those of
typically developing children during reactive balance control, and the resultant control of
balance indicated by the COG-COP relationship. We will describe the differences in balance
control in response to a single support surface perturbation in children with spastic diplegic
cerebral palsy compared to typically developing children of chronologically younger ages
(TDY) and age-matched older children (TDO). The study will investigate the following issues
in balance control: 1) the ability to promptly produce joint torque in response to balance
threat (torque onset latency); 2) the ability to produce large enough magnitude of torque to
counterbalance support-surface perturbation and the relative contributions from the ankle,
knee and hip joints in the recovery of balance (maximum torque); and 3) the speed to reach
peak torque value in response to balance threats (rate of torque production). See Figure 1 for
definition of each variable. In conjunction with our study, we will also discuss other studies
that have contributed to the knowledge of impairments in the musculoskeletal system and
reactive postural control in children with SDCP.
Secondly, we will discuss the culmination of these findings 1) in relation to clinical
applications in the management of musculoskeletal impairments to improve postural control
in children with SDCP, and 2) the significance of using biomechanical measures to show a
direct relationship between the impairments of the musculoskeletal system and reactive
postural control as well as possible coping strategies used by children with SDCP to recover
balance. We will examine the current biomechanical research used to ascertain the
effectiveness of two therapeutic interventions purported to affect the musculoskeletal system
for the improvement of function in children with SDCP: ankle foot orthoses (AFO) and
strength training.
Finally, we will examine the gaps in current clinical research when assessing the
effectiveness of these interventions to reduce the musculoskeletal impairments constraining
static, dynamic, and reactive balance control in children with SDCP.
Figure 1. Schematic for deriving torque baseline, onset latency, maximum torque and rate of torque
generation. Platform movement onset is plotted at 0 in the x-axis.
REACTIVE BALANCE CONTROL IN CHILDREN WITH SDCP

We recruited four children with mild spastic diplegic cerebral palsy (SDCP) aged from 8
to 13 years old for this study. All subjects acquired the disorder prenatally, perinatally or
shortly after birth. Eight typically developing children with no known musculoskeletal or
neurological deficits also participated in the study as control subjects. We formed two
subgroups of the healthy children according to their chronological age: younger typically
developing children (TDY), 4-7 years old, and older typically developing children (TDO), 8-
13 years old, that was age-matched to the children with SDCP. The demographic information
of all participants is listed in Table 1.
The participants stood quietly on a movable dual-forceplate system with arms crossed in
front of the chest. The width of foot placement was self-selected and was approximately
laterally symmetrical with one foot on each forceplate. Perturbations to balance were caused
by simultaneous backward movements of the two forceplates at constant amplitude of 10 cm
and at varying velocities ranging from 5–85 cm/s.
Table 1. Demographic Information of participants
Subject ID Gender GMFCS Age (yr) Weight (kg) Height (m)

SDCP (kg)
1 M III 11.6 22.4 1.24
2 M III 8.5 21.8 1.22
3 M II 12.9 45.5 1.60
4 F II 13.0 27.3 1.36
M 11.50 32.24 1.36
SD 2.10 13.12 0.12
TDY
1 M 7.1 23.6 1.23
2 F 5.2 19.5 1.11
3 F 4.9 14.1 1.04
4 F 6.3 20.2 1.14
M 5.88 19.35 1.23
SD 1.01 3.93 0.08
TDO
1 F 8.3 23.5 1.21
2 F 12.9 34.6 1.41
3 M 11.9 26.1 1.36
4 M 12.5 45.5 1.52
M 11.40 32.43 1.38
SD 2.11 9.92 0.13
We instructed participants not to step unless it was absolutely necessary. Trials were
delivered in three blocks. The perturbation velocity increased by block, but was randomized
within each block (e.g., the platform velocity ranged from 5–25 cm/s in the first block, from
30–55 cm/s in the second block, and from 60–85 cm/s in the last block). We gave catch trials
with no perturbation and with a forward perturbation of 10 cm @ 5 cm/s between blocks to
eliminate anticipatory responses. Catch trials were not included in further analysis.
We obtained kinetics data from the custom-built platform system (Neuroscience
Technical Service Group, University of Oregon, Eugene) consisting of two hydraulically
driven movable forceplates with three axes. Three-dimensional ground reaction forces were
measured by each of the plates independently. For every loaded condition (with participant on
plate), we collected a corresponding unloaded trial (same perturbation characteristics without
participant on plate). We derived net ground reaction forces (ground reaction forces produced
by the participant in response to the perturbation, not including the forces driving the plates)
by subtracting the respective signal in the unloaded trial from that in the loaded trial. Force
data were recorded at 1200 Hz for duration of 5 seconds. Kinematics data were collected
using the PEAK Performance Technologies real-time data-acquisition system (Peak
Performance, Inc., Centennial, CO). Six infrared cameras were positioned around the
participant. We placed reflective markers bilaterally on the following landmarks: (a) the base
of the second metatarsophalangeal joints; (b) lateral malleoli; (c) heels; (d) lateral epicondyles
of the femur; (e) greater trochanters of the femural head; (f), (g), and (h) anterior superior
iliac spines (ASISs) of the pelvis, sacrum, and shoulders, respectively. Reflective wands were
attached on the skin overlying midthighs and midshanks. We placed one reflective marker on
the upper-right corner of the forceplate to monitor the movement of the plates. Before each
experimental session, we took the following anthropometric measures of the participant for
future inverse dynamics analysis: (a) body weight, (b) body height, (c) thigh length, (d)
midthigh circumference, (e) calf length, (f) calf circumference, (g) knee diameter, (h) foot
length, (i) malleolus height, (j) malleolus width, (k) foot breath, and (l) ASIS breadth. Joint
torques were calculated using an inverse dynamics model within the PEAK Performance
Technologies real-time data-acquisition system. We expressed torque as a reaction torque to
all external torques. Counterclockwise torques (view from the right side of the subject) were
expressed in positive values and negative values for clockwise torques.
Previous research in postural control suggests that instead of comparing participants with
differing capabilities under identical perturbation velocities, it is more important to compare
them across conditions that give equivalence of behavioral threat for each individual
(Roncesvalles et al, 2001; Chen and Woollacott, 2007). Thus, for this study, we identified a
set of behaviorally comparable trials to ensure relative equivalence of balance threat across
participants of different sizes, weights and conditions (TDY, TDO and SDCP). As mentioned
earlier, the COG-COP difference is recognized by the balance control system as the error
signal. It is logical to assume that the threat to the balance control system is equivalent when
the sizes of the error signals are similar despite of the different magnitudes of the balance
threats. The comparison condition chosen for this study is threshold velocity. It is defined as
the fastest velocity at which the subject could utilize his/her own muscle contraction and
maintain balance without taking a step.
Due to the uneven force production under the left and right feet for both children with
SDCP and TD children, two force plates were used to record the forces from each of the legs
independently. To account for the asymmetrical weight distribution between the two feet, the
COP used in calculating COG-COP difference was the net COP calculated as follows:
(1)
where
COPl — COP under the left foot
COPr — COP under the right foot
GRFvl — vertical ground reaction force under the left foot
GRFvr — vertical ground reaction force under the right foot.
As shown in Figure 2B, the threshold velocity was significantly different among the three
groups, F(2, 135) = 353.5, p < .001. The TDO children tolerated the highest velocities,
averaging 71.1 ± 0.4 cm/s without taking a step; the TDY children tolerated lower velocities,
51.4 ± 3.0 cm/s; and children with SDCP showed the lowest perturbation tolerance, with
average threshold velocities of 7.9 ± 1.1 cm/s. This indicates that the 8-13 year old children
with SDCP showed even lower developmental level for recovery of postural control to that of
young TD children ranging 4-8 years of age.
Figure 2. A. Normalized center of gravity (COG) and center of pressure (COP) differences at the
threshold velocity. B. Threshold velocity of children in the TDY, TDO and SDCP groups. Error bars
represent standard error of the mean.
As predicted by Winter‘s theory, one-way ANOVA analysis showed that the normalized
(by foot length) COG-COP difference in the anterior-posterior direction at the threshold
velocity was not significantly different among the three groups, indicating that the relative
difficulty of the balance task was equivalent at the threshold velocity, even though the
perturbation velocities were different for each subject (Figure 2A). Therefore, the first part of
the study focused on the analyses of those trials in which subjects experienced the
perturbation at their maximum stability limits.
Torque Onset Latency
One important factor in successfully maintaining stability is activating muscles fast

enough to produce torque in response to external balance threats. To do this, we examined the
joint torque onset latency, which is defined as the time duration between the onset of platform
movement to the time when joint torque exceeded 5 Nm relative to the baseline activity
(calculated from first 50 ms of the trial) and maintain the increase for at least 30 ms. When
we compared the torque onset latencies within each group, the TD children were found to
temporally organize their torque activation pattern; the ankle was activated first, followed by
knee, and the hip was activated last (Figure 3A). A one-way ANOVA showed a significant
difference of onset latency among three joints in both the TDY group, F(2, 33) = 14.5, p <
.001, and the TDO group, F(2, 48) = 113.3, p < .001. On the contrary, children with SDCP
did not show temporally organized activation of joint torques (F(2, 28) = 0.1, p = .868). When
we compared the torque onset latencies between groups, we found a significant difference at
the ankle joint, F(2, 42) = 45.8, p < .001, and the knee joint, F(2, 37) = 10.7, p < .001, but not
the hip joint, F(2, 30) = 0.4, p < .413. Post hoc tests further revealed that at the ankle and
knee joints children with SDCP responded significantly slower than those of the TD children;
however, there was no statistical difference in the ankle torque onset latency between children
in the two TD groups of different ages.
Maximum Torque
The comparison of the maximum torque output (normalized by body weight) produced at
each joint to counterbalance the external perturbation at the threshold velocity (Figure 3B)
showed that children in both TD groups exhibited spatially organized torque contribution
among joints, with the ankle contributing the most and the hip contributing the least. In
children with SDCP, unlike the pattern shown in TD children, the hip joint contributed
significantly more. In fact the maximum hip torque reached a similar level to that of the
ankle. The knee, on the other hand, had a significantly reduced contribution. Post hoc tests
confirmed that the ankle, knee and hip joints were significantly different in the TDO children,
F(2, 48) = 87.6, p < .001; for the TDY children, the pattern is still consistent with an ankle
strategy, where the ankle joint produced a significantly larger torque, there‘s no difference
between the knee and hip joint, F(2, 33) = 11.9, p < .001. In contrast, we found that the ankle
and hip joints contributed equally during balance recovery in children with SDCP; the knee
joint produced significantly lower torque, which potentially compromised the stabilizing
function of the legs.
Figure 3. Joint torque onset latencies (A), normalized maximum torques (B), and normalized rate of
torque productions (C) at the ankle, knee and hip joints at the threshold velocity.
When comparing maximum torque at each joint between groups of all children, we found
no difference between groups at the ankle and knee joints, but the hip torque was significantly
lower in the TDO children and there was no difference between the TDY children and the
children with SDCP. These findings validate previous results using EMG recording and
kinematics measurements indicating that similar to children of chronological younger ages,
children with SDCP use more hip strategy during balance recovery (Sveistrup and
Woollacott, 1996; Roncesvalles et al, 2001; Shumway-Cook and Woollacott, 2001;
Sundermier et al, 2001).
Rate of Peak Torque Production
It has been shown that the rate of torque generation is also critical in balance control
(Robinovitch et al, 2002). A larger maximum torque itself isn‘t sufficient to indicate good
balance, if the speed of reaching the peak value is not fast enough to arrest the acceleration of
the COM. In this study we calculated the rate of joint torque generation by dividing the
maximum joint torque by the duration between the onset of the torque and the time when the
torque reached the maximum value, and then normalized using the individual's body weight.
The results showed that the ankle, knee and hip joints were turned on equally fast within each
group; however, when comparing between groups, it took significantly longer time for
children with SDCP to reach peak torque production at all joints compared to the TD children
(Figure 3C).
In conclusion, it is clear that although the ability to produce torque at ankle and hip joints
are similar in children with SDCP and TD children, maximum torque alone didn‘t allow them
to withstand higher balance threats (see difference in threshold velocity in Figure 2B)
comparable to those of the TD children. There are three main factors contributing to their
impaired balance control. 1) There was a temporal disorganization of joint activation. The TD
children showed clear temporal organization of the torque generation among joints with ankle
being activated first and the hip torque activated the last. The SDCP children, on the other
hand, activated torques at all joints simultaneously; 2) There was a spatial disorganization of
the joint torque profile in children with SDCP. In order to regain balance, children with SDCP
produced equal amount of torques at hip and ankle, lacking the distally focused torque
contribution pattern (ankle produced the largest maximum torque while hip contributed least)
as those found in TD children; and 3) Slow speed to reach peak torque value. It is possible
that spasticity during dynamic motion (spasticity of a muscle being more pronounced when
the joint is in motion) and co-contraction of opposing muscle groups contribute to the slow
rate of torque production, which further impair SDCP children‘s ability to move COP quickly
enough to prevent COG from moving to and beyond the edge of BOS.
An interesting finding in this study is that the normalized maximum joint torques adopted
by children with SDCP at the threshold velocity were no less, if no more, than those seen in
the TD children (see Figure 3B). Wiley and Damiano (1998) documented that children with
spastic cerebral palsy were weaker than age-matched peers in all muscle maximum voluntary
contraction tests. The weakness was more pronounced distally. Moreover, this muscle
weakness was accompanied by muscle imbalance across joints. Hip flexors and ankle plantar
flexors tended to be relatively stronger than their antagonists. Our results in conjunction with
Wiley and Damiano's findings suggest that children with SDCP are likely to use joint torques
near their maximum capacities when recovering from balance threats.
In addition to our research, other studies have also examined the effects of postural
alignment on balance control in children and adults. Burtner et al. (1998) had TD children
stand in the crouched posture alignment, typical of the postural alignment seen in many
children with SDCP, and sustain unexpected platform perturbations. Using EMGs to record
muscle activation patterns in response to recovery of balance, it was shown that 1) ranges of
onset latencies for the TD children in the crouched position were more constrained. 2) The
onset latencies of gastrocnemius and hamstrings muscles were earlier and the frequency of
trunk extensor activation decreased, a trend seen also in children with SDCP. 3) Muscle
organization changed in the crouched position. Distal to proximal organization was seen less
often; and more proximal to distal organization emerged. 4) TD children in the crouched
posture, compared with normal stance, used antagonist muscles more often in response to
platform perturbations. These changes in postural muscle response patterns seen when TD
children were asked to stand in a crouched stance suggest that musculoskeletal constraints are
clear contributors to the atypical postural muscle response patterns seen in children with
SDCP. It also suggests that the neuromuscular response patterns of some children with SDCP
may be appropriate strategies for their musculoskeletal constraints.
A similar study was done by Potter et al. (1990) in which standing balance was assessed
in normal adults while they stood in either 15 or 30 of bilateral knee flexion. The results
indicated that the knee flexion posture increased COP excursion in both the medial-lateral and
anterior- posterior positions. In addition the amount of this increase was linearly affected by
the degree of knee flexion. The greater the degree of the knee flexion angle, the greater the
excursion of COP.
CLINICAL APPLICATIONS
Our study has added to the knowledge of the effects of musculoskeletal constraints on
reactive balance control in children with spastic diplegic cerebral palsy. In addition to muscle
weakness (Wiley and Damiano, 1998) and postural alignment (Burtner et al, 1998; Potter et
al, 1990), the biomechanical measures of lower extremity kinetics in this study demonstrated
that deficits in the temporal and spatial organization of torque production and speed of peak
torque production contribute to the impaired reactive postural control in children with SDCP.
Therefore, therapeutic interventions to improve musculoskeletal impairments that limit the
recovery of balance should not only include techniques and devices to change postural
alignment and muscle force production, but interventions to increase or aid in the rate of
maximal muscle contraction, especially at the ankle during balance recovery. A study by
Engsberg et al. (2006), using 3D gait analysis demonstrated that increasing the strength of the
plantar flexors and dorsi flexors in children with SDCP, alone, did not increase torque
production at the ankle during gait, as was expected.
When making clinical decisions as to the optimal treatment to improve function in
children with cerebral palsy, it‘s important to distinguish the impairments caused by the
primary neural pathophysiology from the impairments caused by coping strategies (Gage,
1991, pg 102). By examining COG-COP difference along with the kinetics of the individual
joints and their temporal/spatial organization, we were able to demonstrate the coping
strategies children with SDCP use to recover balance. Children with SDCP, lacking the
ability to generate torque at the ankle and knee fast enough to recover from perturbations,
generate much greater force at the hip joint (hip strategy) to maintain stability and to recover
balance (compared to what is seen in TD children). The weakness of our study is that the
kinematics at the ankle, knee, and hip joints were not analyzed, because as indicated by
previous studies (Burtner et al, 1998) using EMG recordings, it is possible that a crouched
posture adopted by children with SDCP might account for the disorganized temporal and
spatial torque production at all three joints seen in our study.
AFOs and Strength Training
Next, we will review current research studies using kinematic and kinetic outcomes to
examine the effectiveness of clinical interventions commonly used to correct for postural
alignment and muscle weakness in children with SDCP: ankle foot orthotics (AFOs) and
strength training.
Ankle -foot orthoses (AFO) is one of the most commonly prescribed braces for children
with CP for improving gait by optimizing postural alignment and joint angles. Based on the
understanding of the biomechanics of the human body, it is a brace worn on the lower leg and
foot to support the ankle, hold the foot and ankle in the correct position, and correct foot-
drop. To benefit from the use of AFO, clinical decisions about the management of AFO must
depend on an understanding of the biomechanics of the foot and ankle during gait, the
biomechanics of the orthotic device, and the pathophysiology of gait in children with cerebral
palsy (Davids et al, 2007).
There has been one study examining postural control and the effectiveness of orthotics
(Burtner et al, 1998), by the same research group demonstrated the contribution of
mechanical changes to the altered postural control. In this study the effect of solid and
dynamic ankle-foot orthoses (AFOs) on stance balance control was compared. It was found
that the use of solid AFOs led to decreased activation of distal musculature as well as
decreased recruitment of ankle strategies and increased joint angular velocity at the knee.
Therefore, wearing fixed AFOs would not be advantageous for children with SDCP in the
recovery of balance since the rate of generating sufficient ankle torque is already
compromised. These trends were not found in trials with dynamic AFOs, suggesting that
dynamic AFOs are more advantageous for children with spastic diplegic CP when balance
control is required during unexpected perturbations in stance. This study is particularly
interesting in that it highlighted the importance of using biomechanical-oriented intervention
in the rehabilitation of children with cerebral palsy. It should be of great interest to clinicians
since it emphasizes the impact of alteration in subtle part of the orthoses design on the
efficacy of the treatment.
In the past decade, there has been increasing interest in progressive–resistive strength
training protocols as a therapeutic intervention to improve functional mobility in children
with cerebral palsy (Mockford and Caulton, 2008). Although there have been many 3D gait
analysis (3DGA) studies looking at the effectiveness of strength training on gait, there have
been no studies specifically exploring its effects on static, dynamic, or reactive postural
control. Lowe et al. (2004) demonstrated a strong relationship between isometric muscle force
production and standing balance, but no research studies using the biomechanical measures of
kinematics or kinetics have confirmed this finding.
There are five major attributes of a functional gait which are frequently lost in children
with cerebral palsy: 1) stability in stance, 2) sufficient foot clearance during swing 3) pre
positioning of the foot during terminal swing for initial contact 4) adequate step length, and 5)
energy conservation (Gage, 1991, pg 101). Most clinical research studies examining the
effectiveness of strength training or the use of AFOS to improve gait in children with SDCP
have reported improvements in stride or step length; in the temporal characteristics of
cadence and velocity; and in energy conservation, but little in the improvement of postural
stability (Figueirdo et al, 2008; Mockford and Caulton, 2008).
Though several studies have explored the effects of the different types of AFOs on gait in
children with SDCP (Figueiredo et al, 2008), only a few of these studies have included
dynamic postural control as a dependent variable when assessing the efficacy of the AFOs
using 3D gait analysis technique (Abel et al, 1998; White et al, 2002). By demonstrating an
increase in the percentage of time in single-leg stance versus double-leg stance in the total
gait cycle both studies demonstrated an increase in stability with the use of AFOs when
compared to the barefoot condition. Although these are important measurements to quantify
the increase in stability, without the biomechanical measures of total joint kinetics in relation
to stability, the ‗how‘ of the improvement is not clear (Gage, 1991, pg 23).
Gaps in Clinical Research
Postural control is the underlying motor ability for the development of functional
mobility and higher-level motor skills (Roncesvalles et al, 2001). In addition, postural control
has been shown to change with developmental age and experience in walking in both TD
children and children with SDCP (Burtner et al, 1998). Our study examining reactive postural
control in children with SDCP demonstrated the importance of studying lower extremity
kinetics along with the kinematics and kinetics of whole body movement (COG-COP
relationship) to better understand the mechanics of reactive postural control in children with
SDCP. However, there have been few clinical research studies specifically examining COG-
COP difference as a variable in determining the effectiveness of commonly used interventions
of strength training and AFOs. As demonstrated in our study, knowledge of the whole body
kinematic and kinetic relationship in the recovery of balance, along with the total joint
kinetics has demonstrated the possible coping strategies children with SDCP use to recover
balance when muscle force production is impaired.
Only recently has there been a study that examined dynamic postural control in the gait
of children with SDCP by specifically analyzing COG-COP separation (Hsue et al, 2009).
The separation of COG-COP was measured during an interval of the gait cycle between
single stance of one leg (SL1) to single-leg stance of the opposite leg (SL2) in children with
SDCP and TD of the same age. They found that the RMS of COG-COP in the medial-lateral
(M-L) direction was much greater in children with SDCP compared to the TD children. The
authors suggested the children with SDCP either had a lateral body lean due to weak hip
abductors, or had dynamic postural instability that required a ‗walking by falling‖ pattern of
gait. Further biomechanical measures, i.e., the kinetmatic and kinetic measures of the joints
along with EMGs may reveal the cause of the COG-COP separation in the M-L direction.
Perhaps the children with SDCP spend a greater percentage of their energy expenditure and
generation of joint torgues just to maintain postural stability when walking. Therefore, when
examining the increase in stride length that occurred in children with SDCP as a result of
strength training or the use of AFOs, using biomechanical measures, the question would be 1)
how and if the interventions change the separation of COG-COP and in what direction, 2)
how the total joint kinematics and kinetics change in relation to change in dynamic postural
control, and finally, 3) whether the interventions assist with the compensation strategy to
maintain dynamic postural control, or whether they reduce the musculoskeletal impairments
interfering with postural control.
CONCLUSION
Our study, in addition to studies on the effects of postural alignment on postural control
in children with SDCP, validate the findings of the study by Lowe et al. (2004), in which it is
demonstrated that 54% of the variance in standing postural control (measured by the P-
CTSIB) in children with SDCP could be explained by the limited range of motion in hip,
knee, and ankle joints, as well as the crouch posture; and to abnormal muscle force
production. By analyzing total joint kinetics along with whole body kinematic and kinetics
COG-COP, we demonstrated that children with SDCP use a hip strategy as a coping response
when recovering from external balance threats due to their inadequacy of producing peak
ankle torques fast enough to slow the forward movement of the COM following a
perturbation.
Children with spastic diplegia cerebral palsy are a heterogeneous group, each child
presenting with a variety of musculoskeletal and neuromuscular impairments. Through
3DGA, analyzing the total joint kinetics along with COG-COP separation to measure postural
control, clinicians can better understand the specifics of the coping strategies children with
SDCP use to maintain postural control during gait. From this knowledge, clinical
interventions can be individualized to each child‘s specific impairments, developmental age,
and level of walking experience. These same biomechanical measures could also be used to
examine postural control during other functional activities such as sit-to-stand, crossing
obstacles, walking on uneven terrain and climbing stairs.
ACKNOWLEDGMENTS
The authors wish to thank Marjorie H. Woollacott for her guidance during the authors‘
research at the Motor Control Laboratory of the University of Oregon. The study of reactive
balance control in children with SDCP cited in this chapter was supported by the National
Institutes of Health Grant 2 R01 NS038714-05 to Marjorie H. Woollacott (PI).
REFERENCES
Abel, M.F., Juhl, G.A., Vaughan, C.L., and Damiano, D.L. (1998). Gait assessment of fixed
ankle-foot orthoses in children with spastic diplegia. Arch. Phys. Med. Rehabil., 79(2),
126-33.
Bothner, K. (1998). Postural compensations to a disturbance of balance in human:
Intersegmental dynamics. Diss. Abstr. Internat, 58, 4746.
Burtner, P., Qualls, C., and Woollacott, M. (1998). Muscle activation characteristics of stance
balance control in children with spastic cerebral palsy. Gait Posture, 8, 163-174.
Chen, J., and Woollacott, M. (2007). lower extremity kinetics for balance control in children
with cerebral palsy. J. Motor. Behav., 39(4), 306-316.
Davids, J.R., Rowan, F., and Davis, R.B. (2007). Indications for orthoses to improve gait in
children with cerebral palsy. J. Am. Acad. Orthop. Surg., 15(3), 178-88.
Engsberg, J., Ross, S., and Collins, D. (2006). Increasing ankle strength to improve gait and
function in children with cerebral palsy. Pediatr. Phys. Ther., 18, 266-275.
Figueiredo, E.M., Ferreira, G.B., Moreira, R.C.M., Kirkwood, R.N., and Fetters, L. (2008).
Efficacy of Ankel-Foot Orthoses on Gait of Children with Cerebral Palsy: Systematic
Review of Literature. Pediatr. Phys. Ther., 20, 207-233.
Gage, J.R. Gait Analysis in Cerebral Palsy. London, England: MacKeith Press; 1991.
Horak, F., and Nashner, L. (1986). Central programming of postural movements: Adaptation
to altered support surface configurations. J. Neurophysiol., 55, 1369-1381.
Hsue, B.J., Miller, F., and Su, F.C. (2009). The dynamic balance of the children with cerebral
palsy and typical developing during gait. Part I: Spatial relationship between COM and
COP trajectories. Gait Posture, 29(3), 465-70.
Lowes, L.P., Westcott, S.L., Palisano, R.J., Effgen, S.K., and Orlin, M.N. (2004). Muscle
Force and Range of Motion as Predictors of Standing Balance in Children with Cerebral
Palsy. Phys. Occup. Ther. Pediatr., 24(1-2), 57-77.
Mockford, M., and Caulton, J.M. (2008). Systematic Review of Progressive Strength
Training in Children and Adolescents with Cerebral Palsy Who Are Ambulatory. Pediatr.
Phys. Ther., 20, 318-333.
Nashner, L. (1976). Adaptation reflexes controlling the human posture. Exp. Brain Res., 26,
59-72.
Nashner, L. (1977). Fixed Patterns of rapid postural responses among leg muscles during
stance. Exp. Brain Res., 30, 13-24.
Nashner, L., and McCollum, G. (1985). The organization of human postural movements: A
formal basis and experimental synthesis. Behav. Brain Sci., 8, 135-172.
Nashner, L., Shumway-Cook, A., and Marin, O. (1983). Stance posture control in selected
groups of children with cerebral palsy: Deficits in sensory organization and muscular
coordination. Exp. Brain Res., 49, 393-409.
Potter, P.J., Kirby, R.L., and MacLeod, D.A. (1990). The effects of simulated knee-flexion
contractures on standing balance. Am. J. Phys. Med. Rehabil., 69(3), 144-7.
Robinovitch, S., Heller, B., Lui, A., and Cortez, J. (2002). Effect of strength and speed of
torque development on balance recovery with the ankle strategy. J. Neurophysiol., 88,
613-620.
Roncesvalles, N., Woollacott, M., and Jensen, J. (2001). Development of lower extremity
kinetics for balance control in infants and young children. J. Mot. Behav., 33, 180-192.
Shumway-Cook, A., and Woollacott, M. Motor Control: Theory and practical applications.
Second Edition. Baltimore, Maryland: Lippincott Williams and Wilkins. 2001.
Styer-Acevedo, J. Physical therapy for the child with cerebral palsy. In: Tecklin J, editor.
Pediatric physical therapy. Philadelphia, Pennsylvania: Lippincott Williams and Wilkins;
1994; 89-134.
Sundermier, L., Woollacott, M., Roncesvalles, N., and Jensen, J. (2001). The development of
balance control in children: Comparisons of EMG and kinetics variables and
chronological and developmental groupings. Exp. Brain Res., 136, 340-350.
Sveistrup, H., and Woollacott, H. (1996). Longitudinal development of the automatic postural
response in infants. J. Mot. Behav., 28, 58-70.
Wiley, E., and Damiano, L. (1998). Lower-extremity strength profiles in spastic cerebral
palsy. Dev. Med. Child Neurol., 40, 100-107.
Winter, D. Biomechanics and motor control of human movement. Second Edition. New York,
NY: John Wiley and Sons. 1990.
Winter, D. A.B.C. of balance during standing and walking. Waterloo, Ontario: Wiley. 1995.
White, H., Jenkins, J., Neace, W.P., Tylkowski, C., and Walker, J. (2002). Clinically
prescribed orthoses demonstrate an increase in velocity of gait in children with cerebral
palsy: a retrospective study. Dev. Med. Child Neurol., 44(4), 227-32.
Woollacott, M., Burtner, P., Hensen, J., Jasiewicz, J., Roncesvalles, N., and Sveistrup, H.
(1998). Development of postural responses during standing in healthy children and
children with spastic diplegia. Neurosci. Biobeh. Rev., 22(4), 583-589.
Chapter 8
BIOMECHANICAL PROPERTIES OF CORNEA
Sunil Shah and Mohammad Laiquzzaman

Birmingham Heartland and Solihull NHS Trust, Lode Lane, Solihull,
West Midlands Birmingham, UK, B91 2JL Tel 0121 711 2020 Fax 0121 711 4040
ABSTRACT
The knowledge of corneal biomechanical properties of cornea has gained importance
in recent years. Investigators have been trying to find easy and practical ways to establish
these biomechanical properties but to date have had to rely on corneal thickness measures
to give an idea of corneal biomechanics. This review explores what is known about the
biomechanical properties of the human cornea.
An overview of corneal thickness measurements, its impact on measurement of
intraocular pressure and its importance in various disease states is discussed. The recent
advent of the Ocular Response Analyser, an in-vivo measure of ocular hysteresis and
corneal resistance factor and the pulse waveform associated with this will be discussed.
The importance of this machine with respect to corneal biomechanics will be presented.
INTRODUCTION
The cornea is the transparent outer, fibrous tunic for the eye, which serves a refractive
function while maintaining mechanical strength. These properties, i.e. the refractive power,
mechanical strength and the shape of the cornea, have been a subject of interest to
ophthalmologists for many years. Recently, the importance of the biomechanical properties of
the eye have gained prominence due to the increasing popularity of refractive surgery and the
knowledge of the effects of biomechanics cornea on glaucoma diagnosis and management.
The cornea consists of five layers: epithelium, Bowman‘s membrane, stroma, Descemet‘s
membrane and endothelium. The biomechanical properties of the cornea are a function of all
of the layers, although some structures within the cornea exert a little effect and indeed, there
is no consensus over which layers play a role [1].
Bowman‘s layer has been suggested to play an important role in the biomechanical
stability and rigidity of the cornea [2,3] but other have disputed this suggestion believing that
252 Sunil Shah and Mohammad Laiquzzaman
Bowman‘s layer does not play a significant role. [4,5] The stroma is believed to be main layer
responsible for the corneal biomechanical strength. [6] The stroma constitutes about 90
percent of the corneal thickness, featuring stacked lamellae of collagen fibrils. These collagen
fibrils, which are packed in parallel arrays, make up the 300 to 500 lamellae within the
stroma. The lamellae extend from limbus to limbus and are oriented at various angles to one
another. They are positioned less than 90 degrees to each other in the anterior stroma but
nearly orthogonal in the posterior stroma. [7] In the anterior one third of the stroma, the
collagen lamellae are narrow and thin, undulating, run in random directions and often branch
and interweave in an irregular manner. Studies have postulated that the collagen fibres in the
anterior half of the stroma are the ones responsible for maintaining biomechanical stability
and the corneal shape and hence any surgical incision of these lamellae can result in a loss of
corneal structural stability. [6] However, others have suggested that the collagen fibrils of the
central area are responsible for this stability because they are of higher density and densely
packed. [8] All the collagen lamellae in the posterior two thirds of the stroma are wider and
thicker and tend to be parallel to the corneal surface. The fibrils have a uniform diameter.
[9,10] The lamellae are embedded in a proteo-glycan rich matrix compound of
glycosaminoglycans. [11] This matrix has been suggested to play a role in the absorbance of
mechanical impact and is known to be able to be remodeled.
In endothelial disease, the corneal stroma tends to swell posteriorly more than anteriorly.
However, it has been noted that even extreme swelling of the stroma has a minimal effect on
the curvature of the cornea. [8,12,13] This is probably attributable to the stromal structure.
To date, the only easy method to measure corneal biomechanical properties has been
through the measurement of central corneal thickness (CCT). [14] CCT is used as an
important clinical measurement when assessing corneal health and in the diagnosis of corneal
disease. Therefore, obtaining an accurate CCT measurement is fundamental prior to planning
any surgical intervention of the cornea and assessing the progress of corneal disease.
The Ocular response analyser (ORA) (Name and address of the company)is a new
instrument, which measures corneal biomechanical parameters termed corneal hysteresis
(CH) and corneal resistance factor (CRF) and also measures corneal biomechanically
compensated intraocular pressure (IOPcc). The IOPcc is believed to measure intraocular
pressure (IOP) neutralising the influence of corneal biomechanical properties and CH and
CRF are believed to be the measurement of visco-elastic properties and elasticity of the
cornea respectively.
CENTRAL CORNEAL THICKNESS (CCT)

CCT in humans has been considered a relatively stable parameter in normal eyes. [15]
Doughty and Zaman [16] in their major review after meta-analysis of data from 300
publications reported an average CCT of 534 µm. The range of ‗normal‘ CCT reported in
literature is from 430 µm to 600 µm. [17-19] Although CCT has been found to be at its
highest in the morning just after awakening, it dehydrates and stabilises after 2 to 3 hours and
remains constant throughout the wakening period [20-22].
Measuring CCT is essential to assess the health of the cornea. CCT can be affected by the
diseases processes, various medications and contact lens wear. [23-28] This may result in
Biomechanical Properties of Cornea 253
either increased thickness or thinning of the cornea, which in turn will affect the
biomechanical properties of the cornea.
MEASUREMENT OF CCT
Various instruments have been used to measure CCT, for example, optical pachymetry
[29], non-contact specular microsopy [30], Orbscan II pachymetry [30-34], partial coherence
interferometry [35] and videopachymetry [36].
The increased use of CCT in clinical practice could, however, be attributed to the
availability of easily usable instrumentation i.e. ultrasonic pachymetry. This instrument
provides the gold standard for CCT measurement. [33] Many studies have reported CCT by
ultrasound pachymetry [37].
Ultrasonic CCT is determined by passing a known velocity of ultrasound through the
corneal tissue. The time difference between echoes of signal values reflected from the anterior
and posterior surfaces of the cornea is measured. Typically, high frequency ultrasound
between 10 to 20 Mhz is used to measure CCT. This high working frequency does not allow
echographic scanning of deeper tissue and allow an accuracy within microns. [38,39] The
ultrasound pachymeter requires lower operator skill levels and is portable making it is easy to
use. It has the disadvantage of being an invasive method in which the probe touches the
anterior corneal surface.
CCT AND INTRAOCULAR PRESSURE

Friedenwald [40] recognized the importance of the biomechanical properties of the eye
tunics (ocular rigidity) in the measurement of IOP in his search to develop tonometry. More
recent studies have demonstrated the importance of CCT when assessing the true IOP rather
than measured IOP. [41-47] These studies essentially used the CCT as a measure of corneal
rigidity.
Applanation tonometry (Goldmann Tonometry) has been the gold standard for the
measurement of IOP. It has been assumed that for an applanation area of 3.0 mm to 3.5 mm,
that the capillary attraction of the tear film on the head of tonometer head equals the force
with which the cornea resists flattening. This, however, ignored the effects of surface tension.
Goldmann recognized the importance of corneal rigidity and surface tension as opposing
force. He thus designed his tonometer tip with a diameter of 3.06 mm, which he assumed
would cancel the effect of corneal rigidity and surface tension in an average cornea.
Goldmann [48], however, recognised that IOP measurement could be affected by variation in
CCT in his seminal paper in 1957.
Over a period of many years, a number of studies have considered that CCT can affect
IOP readings. Most of these have expressed an opinion that applanation tonometry readings
will be higher in eyes with thicker CCT and eyes with thinner corneas will display lower IOP
readings. [49-55] This observation confirms that CCT is an important factor and that it can
not be ignored in relation to IOP related diseases and their management.
The relationship between CCT and IOP is, however, not clear. Various factors have been
proposed for this relationship. It is postulated that it may be either due to a measurement error
in applanation tonometry due to differences in corneal thickness, a physiologic effect of IOP
on the corneal tissues, an increase of stromal collagen fibres or corneal stromal rigidity or that
all of these factors may be involved [56].
Studies on patients with diabetes, [57] keratoconus, [58,59] corneal dystrophies and
corneal degenerations and retinal detachment have also shown a relationship between CCT
and IOP [60].
The OHTs study found that ocular hypertensive patients with thicker CCT are at reduced
risk of developing glaucoma as the true IOP was lower. [61-63] Bhatt et al [64] and Manni et
al [65] found that patients with thinner corneas in glaucoma suspects were found more prone
to develop glaucoma i.e. the true IOP was higher. Individual differences in CCT might be
responsible for inaccuracies of the tonometric values with underestimation of the true IOP in
subjects with thinner CCT, although it is becoming clearer that true IOP is unpredictable with
linear correction formulae for CCT and that it is more likely that other biomechanical
properties of the cornea are important. It has been suggested that differences in the corneal
biomechanics may be an expression of individual structural differences of the ocular tissues
with potential consequences on the individual susceptibility to the glaucomatous damage
under the same IOP values. It is important that CCT measurement with other corneal
biomechanical properties should be integrated into the clinical management of glaucoma and
ocular hypertension to better understand the individual risk factors.
It would not be unreasonable to expect that laser ablation of corneal stroma in refractive
surgery will affect IOP. Many studies have reported a lower mean IOP after corneal refractive
surgery. [66-80] Many hypotheses have been presented to explain the cause of lower IOP
readings. It has been postulated that as the collagen fibres are ablated during the refractive
surgery with the cut collagen fibres losing their tensile strength. [81] Others have suggested
that the general softening of the corneal tissue, following the natural healing process, may be
a responsible factor [82] while others suggest that decrease in CCT and change of corneal
curvature might be the major cause. [83]
However, without being able to measure these biomechanical properties in vivo, these
influences cannot be either predicted or controlled [84].
CORNEAL BIOMECHANICS
The elasticity of a given tissue or material is described as the relationship between stress
and strain. Stress is described as the force per unit cross section applied on a tissue material.
Strain measures the stretch of a material and is calculated as a change in the length of the
tissue divided by its original length. The elastic modulus, or Youngs Modulus (Y) is the ratio
of the change in stress to the associated change in strain. Little deformation occurs with an
increase in strain when the elastic modulus is high (Y). The smaller the Y, the more elastic
the tissue [85-86].
However, for biological material, Y is not constant. The visco-elastic response consists of
an immediate deformation followed by a slow deformation. [86] Hence, the cornea reacts to
stress as a typical visco-elastic material [87].
The maintenance of corneal shape is determined by IOP, the elastic properties of the
corneal tissue and amount and distribution of the corneal tissue mass. [88] The biomechanical
properties of the eye, in the past, has been investigated extensively, but most of these studies
have been performed in-vitro (either on whole eye or excised corneal tissue). These studies
include measuring the tensile strength of strips the corneal tissue isolated by dissection and
loaded with weights to cause extension and monitoring changes in the spacing between
cellular and extracellular particles under load. [85-96] Recently some studies have reported
elasticity of eyes in vivo by injecting saline solution in to the anterior chamber [97], by
measuring CH and CRF by ORA [98] and by dynamic corneal imaging [99].
Edmund [86] investigated the rigidity of the cornea by measuring the radius of the central
corneal curvature, the coefficient of radius variation, the CCT and the coefficient of thickness
variation. Friedenwald [40], devised a formula for ocular rigidity where K (an assumed
constant scleral rigidity) = dP/dV, (V is intraocular volume and P is the IOP). He assumed
that K is constant but later studies have found that K is not constant and that it was found to
decrease with IOP in human eyes. Investigators have concluded that the phenomenon of
ocular rigidity was complex, and even with very simplifying assumptions, the analysis did not
lead to a simple result.
Orssengo and Pye, [100] more recently, determined the modulus of corneal rigidity
(elasticity) in vivo from corneal dimensions and applanation tonometry.
Although all these above mentioned studies are very interesting and have tried to
establish ocular rigidity by different methods. They all involve complicated mathematical
calculations [86,88,90,92] and are impractical for clinicians and (also report varying results).
MEASUREMENT OF BIOMECHANICAL PROPERTIES OF THE CORNEA

An instrument that provides in vivo measurement values and which is easy to use in a
clinical setting, easier for clinicians in terms of data interpretation, sufficiently reliable
(accurate) and is reproducible will be ideal for the purpose of measuring biomechanical
properties. The first such instrument is the ORA.
The ORA determines the biomechanical properties of the cornea by using a bi-directional
applanation process. This is achieved using a force-displacement relationship where the
instrument releases a precisely metered air pulse, which strikes the central cornea causing it to
move inward. This is followed by flattening of the cornea, which then finally becomes
concave. The flattening of the corneal surface is termed ‗first applanation‘ or ‗inward
applanation‘. A millionth of a second after inward applanation, the air pulse is shut down,
resulting in a decrease in the pressure applied to the cornea. As the pressure decreases, the
cornea returns from the concave state to flatten again (second applanation or outward
applanation) and then finally resumes its normal convex shape. The change in the corneal
curvature is monitored by an electro-optical collimation detector in the central 3.0 mm of the
cornea and the whole process lasts just 20 milliseconds.
A filtered version of the detector signal defines the two applanations as two peaks
(inward and outward applanation). The two pressures of internal air supply plenum are
determined from the applanation times derived from the detector applanation peaks. The
inward and outward applanation pressures are different because of the visco-elastic properties
of the cornea. However, theoretical calculations indicate that area density of the cornea is a
minor contributing factor to the pressure difference. These are perhaps the result of immediate
elastic response and delayed or steady state elastic responses respectively of the corneal
tissue. Energy adsorption during the corneal deformation delays the occurrence of the
applanation signal peaks resulting in a difference between the applation pressures. This
difference between the two peaks is referred as corneal hysteresis (CH). CH is believed to
reflect the aggregate effects of thickness, rigidity, hydration and perhaps other factors yet to
be determined. Corneal resistance factor (CRF) is believed to be a measure of the elastic
property of the cornea [98].
NORMAL CORNEAL HYSTERESIS (CH) AND CORNEAL RESISTANCE

FACTOR (CRF)
The ORA has been reported to give reproducible measurements of these parameters.
[101,102] CH and CRF are constant and these parameters do not have a diurnal variation
during the day time. [103-105] No correlation has been found between age, gender, corneal
steepness, axial length and spherical equivalent refraction do not affect CH and CRF values
[106-108].
The mean CH in normal eyes in various publications so far has been reported to be
between 10.1 mm Hg and 12. 7 mm Hg (range 4.3 mm Hg to 16.7 mm Hg) and mean CRF
values between 10 mm Hg and 11. 7 mm Hg (range 4.0 mm Hg to 17.1 mm Hg). [109-111]
There appears to be no agreement on the CH measurement in children, some have found no
difference compared to adults while others have found CH and CRF values in children to be
slightly higher and others have found lower values [112,113].
As one would expect, there should be a relationship between CCT and CH and CRF, as
thicker corneas should be more rigid due to the increased mass of corneal tissue. Studies have
reported these parameters to be positively correlated i.e. the higher the CCT the higher CH
and CRF. [106,109] However, these reported studies have so far revealed that CH and CRF
are independent biomechanical parameters from CCT. In addition, studies have found a
moderate to significant dependence of CH and CRF on IOP and CCT [106,114].
ORA READINGS IN DISEASED EYES

CH and CRF in eyes with glaucoma has been previously investigated. [115-118] Shah et
al (2008) reported POAG and normal tension glaucoma eyes (NTG) to have lower mean CH
values than OHT eyes [117]. Mean CH in NTG, POAG and OHT eyes was 9.0 +/- 1.9, 9.9
and 10.2 mm Hg; CRF was 9.1, 10.6 and 12.0 mm Hg respectively [117].
Weaker corneas could be screened with ORA parameters, and low CH could be
considered a risk factor for underestimation of IOP, however, this does not mean to reduce
the effect of CCT on IOP and CCT should continue to be considered a useful parameter
[114].
Studies on the keratoconic eyes revealed that CH and CRF show a reduced value
compared to normal eyes. CH has been reported to be found between 7.5 mm Hg to 9.6 mm
Hg [106,114,119,120].
Several studies have reported on the comparison of CH and CRF in keratoconus eyes
with LASIK eyes. [114,120,121] These studies compared CH an CRF of post-surgical eyes
with those of keratoconus eyes and found that these parameters in keratoconus eyes were
lower than in post-LASIK eyes but Shah and Laiquzzaman [121] found that these
biomechanical parameters were very similar when comparing keratoconic and post-refractive
surgery eyes. In the study by Shah and Laiquzaman [121] post-refractive surgery mean CH
was 9.2 mmHg and CRF 7.6 mmHg. The values for keratoconic eyes were 9.4 mmHg and 7.7
mmHg respectively.
Luce [98] reported that applanation signal morphology in post-refractive surgery eyes
indicated similar pattern that of keratoconus and Fuch‘s dystrophy eyes indicating reduced or
low biomechanical properties in these eyes.
REFRACTIVE SURGERY AND CORNEAL BIOMECHANICAL

PROPERTIES
Corneal refractive surgery results in a reduction of CCT and a change in corneal
curvature. Understanding the link between the microstructure of cornea and its biomechanics
is fundamental in eyes undergoing laser eye surgery.
There has been a general agreement that individual variation in the biomechanical
properties of the cornea all influence treatment outcome and that a method for pre-operative
assessment of corneal biomechanical properties of individual eye will may help improve
outcome.
To predict refractive surgery outcome, ablation algorithms are derived from the statistical
analyses of cumulative databases. These algorithms do not deal with corneal biomechanical
measurements and their individual variations, because these formulae have been calculated
without an understanding of the physical response of the cornea to structural change from
ablation of the corneal stroma. A basic understanding of intra-operative corneal
biomechanical changes is necessary to improve predictability, safety and effectiveness in the
treatment [122].
Laser refractive surgery involves flap creation and excimer laser ablation, which may
affect and alter the biomechanics of the cornea. Many investigators have reported a mean
reduction of CH and CRF after laser refractive surgical treatment. Ortiz et al [120] reported
that CH and CRF were significantly reduced in eyes after LASIK. Pepose et al [123] also
reported a decreased CH and CRF after laser refractive surgery. They found that CH
decreased by 16.2% and CRF decreased by a mean 28.6%. Other studies have found the same
results where the CRF values were found to be reduced more than CH. [120,121,123-125]
The results of these studies imply that changes in CH and CRF after refractive surgery
indicate an alteration in corneal biomechanics correlating with the attempted correction. Chen
et al [125] conclude that CRF may be more useful than the CH in assessing biomechanical
changes resulting from LASIK.
Refractive surgery carries a risk of inducing iatrogenic keratectasia, although this

incidence is low 0.66%. [126] 250 µm has been regarded as arbitrary safe minimum CCT
after LASIK. Other corneal surgery protocols have recommend that the total corneal
thickness at the end of the procedure should be greater than 400 µm (in order that the residual
stroma does not incur ectasia) [127]. According to the suggested protocol, eyes with a total
CCT less than 400 µm should be considered at risk for ectasia. [128] Keratoectasia is more
common in eyes with residual thin beds and high IOP. [129] Having said this, keratoectasia
can also be observed in small corrections. [130,131] These observations imply that there must
be differences in individual corneal biomechanical properties.
THE FUTURE
Several other instruments, which can be useful in the future to know biomechanical
properties of the cornea have also been reported in the literature.
The holographic interferometry technique can determine the biomechanical properties in
vivo. [132,133] This technique has been reported to measure the elasticity of the cornea. It is a
non-destructive optical comparison technique analogous to corneal topography, in that fringes
are generated from the surface displacements of the cornea and then recorded. [134] Under
stress, the changes in the induced strain are seen as changes in the distribution and spacing of
the fringes. The high sensitivity of this method enables measurement of the radial distension
of the cornea even for very small IOP changes equal to 0.0375 mm Hg. This technique has
practical limitations. The holographic technique generates whole-field data across the cornea.
This vast data is manually assessed and then interpreted with simple cross-sectional graphs of
displacement. This is impractical in a busy clinical setup.
Another technique reported to measure biomechanical properties is non-contact
electronic speckle pattern interferometry (ESPI). This instrument also measures the elasticity
of the cornea. ESPI does not require photographic recording of a hologram thus simplifying
the measurements and provides rapid real-time measurement of surface displacement. An
ESPI has the ability to monitor changes in biological tissue under stress with sub-micron
accuracy and to acquire such information rapidly and in a non-destructive manner. The
benefits of using the ESPI is that it is non-contact, the measurement can be made of the whole
cornea, it has the sensitivity to detect and monitor corneal displacements generated in vivo
and digital processing allows the real-time measurements and measurements of displacement
up to 0.01 µm [122].
Grabner et al [99] reported the measurement of corneal elasticity using dynamic corneal
imaging in vivo using computer assisted veideokeratography. This is performed using
dynamic corneal imaging method by central identation of the cornea to assess the elastic
properties of the eye.
The concept of ultrasound elasticity microscope is also a further step towards
understanding the elastic properties of cornea. It produces high-quality conventional
ultrasonic B-scans over large thicknesses by confocal processing. The ultrasound elasticity
microscope can produce high resolution strain images throughout the corneal depth and
various corneal layers with different elastic properties appearing as different strains in the
images.
Development of these new instruments may potentially lead to an increased

understanding of corneal biomechanics.
REFERENCES
[1] Mow CC. A theoretical model of the cornea for use in studies of tonometry. Bull Math
Biophys. 1968; 30: 437-53.
[2] Leibowitz HM. Corneal Disorders: Clinical diagnosis and management. 1984.
Saunders, Philadelphia (p14).
[3] Morishige N, Petroll WM, Nishida T, Kenney MC, Jester JV. Noninvasive corneal
stromal collagen imaging using two-photon-generated second-harmonic signals. J.
Cataract. Refract. Surg. 2006; 32: 1784-91.
[4] Montés-Micó R, Charman WN. Intraocular pressure after excimer laser myopic
refractive surgery. Ophthalmic. Physiol. Opt. 2001; 21: 228-35.
[5] Hjortdal JO, Ehlers N. Effect of excimer laser keratectomy on the mechanical
performance of the human cornea. Acta. Ophthalmol. Scand. 1995; 73: 18-24.
[6] Boote C, Dennis S, Newton RH, Puri H, Meek KM. Colleagen fibrils appear more
closely packed in the prepupillary cornea: optical and biomechanical implications.
Invest. Ophthalmol. Vis. Sci. 2003; 44: 2941-8.
[7] Klyce SD, BeuermanR. Structure and Function of the Cornea In The Cornea(Eds,
Kaufman, H., Barron, B. and McDonald, M. B.) 1997 Butterworth-Heinemann, Boston,
pp. 3-50.
[8] Müller LJ, Pels E, Vrensen GF. The specific architecture of the anterior stroma
accounts for maintenance of corneal curvature. Br. J. Ophthalmol. 2001; 85: 437-43.
[9] Pouliquen Y J. 1984 Castroviejo lecture. Fine structure of the corneal stroma. Cornea.
1984-85; 3: 168-77.
[10] Komai Y, Ushiki T. The three dimensional organisation of collagen fibrils in the human
cornea and sclera. Invest. Ophthalmol. Vis. Sci. 1991; 32: 2244-58.
[11] Davison PF, Galbavy EJ. Connective tissue remodeling in corneal and scleral wounds.
[12] Erickson P, Comstock TL, Doughty MJ, Cullen AP. The cornea swells in the posterior
direction under hydrogel contact lenses Ophthalmic. Physiol. Opt. 1999; 19: 475-80.
[13] Moezzi AM, Fonn D, Simpson TL, Sorbara L. Contact lens-induced corneal swelling
and surface changes measured with the Orbscan II corneal topographer. 2004. Optom.
Vis. Sci; 81: 189-93.
[14] Brubaker RF. Tonometry and corneal thickness. Arch Ophthalmol. 1999; 117: 104-5.
[15] Mishima S. Corneal thickness. Surv. Ophthalmol. 1968; 13: 57-96.
[16] Doughty MJ, Zaman ML. Human corneal thickness and its impact on intraocular
pressure: a review and meta-analysis approach. Surv. Ophthalmol. 2000; 44: 367-408.
[17] Alsbirk PH. Corneal thickness. I. Age variation, sex difference and oculometric
correlations. Acta. Ophthalmol. 1978; 56: 95-104.
[18] Alsbirk PH. Corneal thickness. II. Environmental and genetic factors. Acta.
Ophthalmol. 1978; 56: 105-13.
[19] Avitabile T, Marano F, Uva MG, Reibaldi A. Evaluation ocentral and peripheral
corneal thickness with ultrasound biomicroscopy in normal and keratoconus eyes.
Cornea. 1997; 16: 639-44.
[20] Manchester PT Jr. Hydration of the cornea. Trans. Am. Ophthalmol. Soc. 1970; 68:
425-61.
[21] Kiely PM, Carney LG, Smith G. Diurnal variations of corneal topography and
thickness. Am. J. Optom. Physiol. Opt. 1982; 59: 976-82.
[22] Harper CL, Boulton ME, Bennett D, Marcyniuk B, Jarvis-Evans JH, Tullo AB,
Ridgway AE. Diurnal variations in human corneal thickness. Br. J. Ophthalmol. 1996;
80: 1068-72.
[23] Ehlers N, Bramsen T.Central thickness in corneal disorders. Acta. Ophthalmol. 1978;
56: 412-6.
[24] Bramsen T, Klauber A, Bjerre P. Central corneal thickness and intraocular tension in
patients with acromegaly. Acta. Ophthalmol. 1980; 58: 971-4.
[25] Katz B, Melles RB, Schneider JA, Rao NA.Corneal thickness in nephropathic
cystinosis. Br. J. Ophthalmol. 1989; 73: 665-8.
[26] Yaylali V, Kaufman SC, Thompson HW.Corneal thickness measurements with the
Orbscan Topography System and ultrasonic pachymetry. J. Cataract. Refract. Surg.
1997; 23: 1345-50.
[27] Høvding G.The central corneal thickness in keratoconjunctivitis sicca. Acta.
Ophthalmol. 1992; 70: 108-10.
[28] Liu Z, Pflugfelder SC. Corneal thickness is reduced in dry eye. Cornea. 1999; 18: 403-
7.
[29] Hansen FK. A clinical study of the normal human central corneal thickness. Acta.
Ophthalmol. 1971; 49: 82-8.
[30] Olsen T, Ehlers N. The thickness of the human cornea as determined by a specular
method. Acta. Ophthalmol. (Copenh). 1984; 62: 859-71.
[31] Liu Z, Huang AJ, Pflugfelder SC. Evaluation of corneal thickness and topography in
normal eyes using the Orbscan corneal topography system. Br. J. Ophthalmol. 1999;
83: 774-8.
[32] Kawana K, Tokunaga T, Miyata K, Okamoto F, Kiuchi T, Oshika T. Comparison of
corneal thickness measurements using Orbscan II, non-contact specular microscopy,
and ultrasonic pachymetry in eyes after laser in situ keratomileusis. Br. J. Ophthalmol.
2004; 88: 466-8.
[33] Gherghel D, Hosking SL, Mantry S, Banerjee S, Naroo SA, Shah S. Corneal
pachymetry in normal and keratoconic eyes: OrbscanII versus ultrasound. J. Cataract.
Refract. Surg. 2004; 30:1272-7.
[34] Rainer G, Findl O, Petternel V, Kiss B, Drexler W, Skorpik C, Georgopoulos M,
Schmetterer L. Central corneal thickness measurements with partial coherence
interferometry, ultrasound, and the Orbscan system. Ophthalmology. 2004; 111: 875-9
[35] Hitzenberger CK, Baumgartner A, Drexler W, Fercher AF. Interferometric
measurement of corneal thickness with micrometer precision. Am. J. Ophthalmol. 1994;
118: 468-76.
[36] Franco S, Almeida JB, Parafita M. Measurement of corneal thickness by
videopachymetry: preliminary results. J. Refract Surg. 2000; 16: S661-3.
[37] McLaren JW, Bourne WM. A new videopachometer. Invest. Ophthlamol. Vis. Sci.
1999; 40: 1593-8.
[38] Ling T, Ho A, Holden BA. Method of evaluating ultrasonic pachometers. Am. J. Optom
Phys Optics. 1986; 63:462-66.
[39] Sabetti L, Spadea L, Furcese N, Balestrazzi E. Measurement of corneal thickness by
ultrasound after photorefractive keratectomy in high myopia. J. Refract. Corneal. Surg.
1994; 10(2 Suppl): S211-6.
[40] Friedenwald JS. Contribution to the theory and practice of tonometry. Am. J.
Ophthalmol. 1937; 20: 985-1024.
[41] Whitacre MM, Stein RA, Hassanein K. The effect of corneal thickness on applanation
tonometry. Am. J. Ophthalmol. 1993; 115: 592-6.
[42] Herndon LW, Choudhri SA, Cox T, Damji KF, Shields MB, Allingham RR. Central
corneal thickness in normal, glaucomatous, and ocular hypertensive eyes. Arch.
Ophthalmol. 1997; 115: 1137-41.
[43] Foster PJ, Baasanhu J, Alsbirk PH, Munkhbayar D, Uranchimeg D, Johnson GJ.
Central corneal thickness and intraocular pressure in a Mongolian population.
Ophthalmology. 1998; 105: 969-73.
[44] Stodtmeister R. Applanation tonometry and correction according to corneal thickness.
Acta. Ophthalmol. 1998; 76: 319-24.
[45] Tanaka GH. Corneal pachymetry: a prerequisite for applanation tonometry? Arch.
Ophthalmol. 1998; 116: 544-5.
[46] Shah S, Chatterjee A, Mathai M, Kelly SP, Kwartz J, Henson D, McLeod D.
Relationship between corneal thickness and measured intraocular pressure in a general
ophthalmology clinic. Ophthalmology. 1999; 106: 2154-60.
[47] Aghaian E, Choe JE, Lin S, Stamper RL. Central corneal thickness of Caucasians,
Chinese, Hispanics, Filipinos, African Americans, and Japanese in a glaucoma clinic.
Ophthalmology. 2004; 111: 2211-9.
[48] Goldmann H, Schmidt T. Applanation tonometry. Ophthalmologica. 1957; 134: 221-4.
[49] Tomlinson A, Leighton DA. Ocular dimensions in low tension glaucoma compared
with open-angle glaucoma and the normal. Br. J. Ophthalmol. 1972; 56: 97-105.
[50] Ehlers N, Hansen FK. Central corneal thickness in low-tension glaucoma. Acta.
Ophthalmol.. 1974; 52: 740-6.
[51] Johnson M, Kass MA, Moses RA, Grodzki WJ. Increased corneal thickness simulating
elevated intraocular pressure. Arch. Ophthalmol. 1978; 96: 664-5.
[52] Swan PG, Brown B, Wlidsoet CF. The influence of sleep and central corneal thickness
on intraocular pressure in a case of glaucoma. Practical Optom. 1990; 1: 112-5.
[53] Argus WA. Ocular hypertension and central corneal thickness. Ophthalmology. 1995;
102: 1810-2.
[54] Herman DC, Hodge DO, Bourne WM. Increased corneal thickness in patients with
ocular hypertension. Arch. Ophthalmol. 2001; 119: 334-6.
[55] Bhan A, Browning AC, Shah S, Hamilton R, Dave D, Dua HS. Effect of corneal
thickness on intraocular pressure measurements with the pneumotonometer, Goldmann
applanation tonometer, and Tono-Pen. Invest. Ophthalmol. Vis. Sci. 2002; 43: 1389-9.
[56] Wolfs RC, Klaver CC, Vingerling JR, Grobbee DE, Hofman A, de Jong PT.
Distribution of central corneal thickness and its association with intraocular pressure:
The Rotterdam Study. Am. J. Ophthalmol. 1997; 123: 767-72.
[57] Skaff A, Cullen AP, Doughty MJ, Fonn D. Corneal swelling and recovery following
wear of thick hydrogel contact lenses in insulin-dependent diabetics. Ophthalmic.
Physiol. Opt. 1995; 15: 287-97.
[58] Böhm A, Kohlhaas M, Lerche RC, Bischoff B, Richard G. Measuring intraocular
pressure in keratoconus. Effect of the changed biomechanics. Ophthalmologe. 1997; 94:
771-4.
[59] Patel S, McLaughlin JM. Effects of central corneal thickness on measurement of intra-
ocular pressure in keratoconus and post-keratoplasty. Ophthalmic. Physiol. Opt. 1999;
19: 236-41.
[60] Winther J. Central corneal thickness in patients with retinal detachment. Acta.
Ophthalmol. 1982; 60: 945-8.
[61] Brandt JD, Beiser JA, Kass MA, Gordon MO. Central corneal thickness in the Ocular
Hypertension Treatment Study (OHTS). Ophthalmology. 2001; 108: 1779-8.
[62] Gordon MO, Beiser JA, Brandt JD, Heuer DK, Higginbotham EJ, Johnson CA, Keltner
JL, Miller JP, Parrish RK 2nd, Wilson MR, Kass MA. The Ocular Hypertension
Treatment Study: baseline factors that predict the onset of primary open-angle
glaucoma. Arch. Ophthalmol. 2002; 120: 714-20.
[63] Brandt JD. Central corneal thickness, tonometry, and glaucoma risk--a guide for the
perplexed. Can. J. Ophthalmol. 2007; 42: 562-6.
[64] Bhatt N, Bhojwani R, Morrison A, Kwartz J, Laiquzzaman M, Shah S. A 10-year
follow up of ocular hypertensive patients within the Bolton Corneal Thickness Study.
Can measured factors predict prognostic outcomes?Cont. Lens. Anterior. Eye. 2008; 31:
147-53.
[65] Manni G, Oddone F, Parisi V, Tosto A, Centofanti M. Intraocular pressure and central
corneal thickness. Prog. Brain Res. 2008; 173: 25-30.
[66] Kohlhaas M, Lercher R-CH, Draeger J, Klemm M, Ehlers N, Hjordtal J. Olsen H,
Barraquer C, Barraquer JI, Flicker D, Rivera F, Carriazo C. The influence of central
corneal thickness and corneal curvature after corneal refractive surgery. Eur. J. Implant.
Ref. Surg. 1995; 7: 84-8.
[67] Chatterjee A, Shah S, Bessant DA, Naroo SA, Doyle SJ. Reduction in intraocular
pressure after excimer laser photorefractive keratectomy. Correlation with pretreatment
myopia. Ophthalmology. 1997; 104: 355-9.
[68] Mardelli PG, Piebenga LW, Whitacre MM, Siegmund KD. The effect of excimer laser
photorefractive keratectomy on intraocular pressure measurements using the Goldmann
applanation tonometer. Ophthalmology. 1997; 104: 945-8.
[69] Schipper I, Senn P, Oyo-Szerenyi K, Peter R. Central and peripheral pressure
measurements with the Goldmann tonometer and Tono-Pen after photorefractive
keratectomy for myopia. J. Cataract. Refract. Surg. 2000; 26: 929-33.
[70] Rashad KM, Bahnassy AA. Changes in intraocular pressure after laser in situ
keratomileusis. J. Refract. Surg. 2001; 17: 420-7.
[71] Oliveira-Soto L, Charman WN. Some possible longer-term ocular changes following
excimer laser refractive surgery. Ophthalmic. Physiol. Opt. 2002; 22: 274-88.
[72] Zadok D, Raifkup F, Landao D, Frucht-Pery J. Intraocular pressure after LASIK for
hyperopia. Ophthalmology. 2002; 109: 1659-61.
[73] Kaufmann C, Bachmann LM, Thiel MA. Intraocular pressure measurements using
dynamic contour tonometry after laser in situ keratomileusis. Invest. Ophthalmol. Vis.
Sci. 2003; 44: 3790-4.
[74] Cheng AC, Leung DY, Cheung EY, Fan DS, Law RW, Lam DS. Intraocular pressure
measurement in patients with previous LASIK surgery using pressure phosphene
tonometer. Clin. Experiment Ophthalmol. 2005; 33: 153-7.
[75] Chihara E, Takahashi H, Okazaki K, Park M, Tanito M. The preoperative intraocular
pressure level predicts the amount of underestimated intraocular pressure after LASIK
for myopia. Br. J. Ophthalmol. 2005; 89: 160-4.
[76] Jarade EF, Abi Nader FC, Tabbara KF. Intraocular pressure measurement after
hyperopic and myopic LASIK. J Refract Surg. 2005; 21: 408-10.
[77] Hjortdal JØ, Møller-Pedersen T, Ivarsen A, Ehlers N. Corneal power, thickness, and
stiffness: results of a prospective randomized controlled trial of PRK and LASIK for
myopia. J. Cataract. Refract. Surg. 2005; 31: 21-9.
[78] Siganos DS, Papastergiou GI, Moedas C. Assessment of the Pascal Dynamic contour
tonometer in monitoring intraocular pressure in unoperated eyes and operated eyes after
LASIK. J. Cataract. Refract. Surg. 2005; 31: 458-459.
[79] Svedberg H, Chen E, Hamberg-Nyström H. Changes in corneal thickness and curvature
after different excimer laser photorefractive procedures and their impact on intraocular
pressure measurements. Graefes Arch. Clin. Exp. Ophthalmol. 2005; 243: 1218-20.
[80] Hsu SY, Chang MS, Lee CJ. Intraocular pressure assessment in both eyes of the same
patient after laser in situ keratomileusis. J. Cataract. Refract. Surg. 2009; 35: 76-82.
[81] Munger R, Dohadwala AA, Hodge WG, Jackson WB, Mintsioulis G, Damji KF.
Changes in measured intraocular pressure after hyperopic photorefractive keratectomy.
J. Cataract. Refract. Surg. 2001; 27: 1254-62.
[82] Patel S, Aslanides IM. Main causes of reduced intraocular pressure after excimer laser
photorefractive keratectomy. J. Refract. Surg. 1996; 12: 673-4.
[83] Montés-Micó R, Charman WN. Intraocular pressure after excimer laser myopic
refractive surgery. Ophthalmic. Physiol. Opt. 2001; 21: 228-35.
[84] Roberts C. Biomechanical customization: the next generation of laser refractive
surgery. J. Cataract. Refract. Surg. 2005; 31: 2-5.
[85] Battaglioli JL, Kamm RD. Measurements of the compressive properties of scleral tissue
[86] Edmund C. Corneal elasticity and ocular rigidity in normal and keratoconic eyes. Acta.
Ophthalmol. 1988; 66: 134-40.
[87] Nash IS, Greene PR, Foster CS. Comparison of mechanical properties of keratoconus
and normal corneas. Exp. Eye Res. 1982; 35: 413-24.
[88] Edmund C. Assessment of an elastic model in the pathogenesis of keratoconus. Acta.
Ophthalmol. 1987; 65: 545-5.
[89] Hartstein J, Becker B. Research into the pathogenesis of keratoconus. A new syndrome:
low ocular rigidity, contact lenses, and keratoconus. Arch. Ophthalmol. 1970; 84: 728-
9.
[90] Foster CS, Yamamoto GK Ocular rigidity in keratoconus. Am. J. Ophthalmol. 1978; 86:
802-6.
[91] Andreassen TT, Simonsen AH, Oxlund H. Biomechanical properties of keratoconus
and normal corneas. Exp. Eye Res. 1980; 31: 435-4.
[92] Brooks AMV, Robertson IF, Mahoney A. Ocular rigidity and intraocular pressure in
keratoconus. Aust. J. Ophthalmol. 1984; 12: 317-24.
[93] Reichel E, Miller D, Blanco E, Mastanduno R. The elastic modulus of central and
perilimbal bovine cornea. Ann. Ophthalmol. 1989; 21: 205-8.
[94] Vito RP, Carnell PH. Finite element based mechanical models of the cornea for
pressure and indenter loading. Refract. Corneal. Surg. 1992; 8: 146-51.
[95] Hjortdal JO, Jensen PK. In vitro measurement of corneal strain, thickness, and
curvature using digital image processing. Acta. Ophthalmol. Scand. 1995; 73: 5-11.
[96] Purslow PP, Karwatowski WSS. Ocular elasticity; Is engineering stiffness a more
useful characterization parameter than ocular rigidity? Ophthalmology. 1996; 103:
1686-92.
[97] Pallikaris IG, Kymionis GD, Ginis HS, Kounis GA, Tsilimbaris MK. Ocular rigidity in
living human eyes. Invest. Ophthalmol. Vis. Sci. 2005; 46: 409-14.
[98] Luce DA. Determining in vivo biomechanical properties of the cornea with an ocular
response analyzer. J. Cataract. Refract. Surg. 2005; 31: 156-62.
[99] Grabner G, Eilmsteiner R, Steindl C, Ruckhofer J, Mattioli R, Husinsky W. Dynamic
corneal imaging. J. Cataract. Refract. Surg. 2005; 31: 163-74.
[100] Orssengo GJ, Pye DC. Determination of the true intraocular pressure and modulus of
elasticity of the human cornea in vivo. Bull Math Biol. 1999; 61: 551-72.
[101] Kynigopoulos M, Schlote T, Kotecha A, Tzamalis A, Pajic B, Haefliger I. Repeatability
of intraocular pressure and corneal biomechanical properties measurements by the
ocular response analyser. Klin. Monatsbl. Augenheilkd. 2008; 225: 357-60.
[102] Moreno-Montañés J, Maldonado MJ, García N, Mendiluce L, García-Gómez PJ, Seguí-
Gómez M. Reproducibility and clinical relevance of the ocular response analyzer in
nonoperated eyes: corneal biomechanical and tonometric implications. Invest.
Ophthalmol. Vis. Sci. 2008; 49: 968-74.
[103] Laiquzzaman M, Bhojwani R, Cunliffe I, Shah S. Diurnal variation of ocular hysteresis
in normal subjects: relevance in clinical context. Clin. Exp. Ophthalmol. 2006; 34: 114-
8.
[104] González-Méijome JM, Queirós A, Jorge J, Díaz-Rey A, Parafita MA. Intraoffice
variability of corneal biomechanical parameters and intraocular pressure (IOP). Optom.
Vis Sci. 2008; 85: 457-62.
[105] Shen M, Wang J, Qu J, Xu S, Wang X, Fang H, Lu F. Diurnal variation of ocular
hysteresis, corneal thickness, and intraocular pressure. Optom. Vis Sci. 2008; 85: 1185-
92.
[106] Shah S, Laiquzzaman M, Bhojwani R, Mantry S, Cunliffe I. Assessment of the
biomechanical properties of the cornea with the ocular response analyzer in normal and
keratoconic eyes. Invest. Ophthalmol. Vis. Sci. 2007; 48: 3026-31.
[107] Kida T, Liu JH, Weinreb RN. Effects of aging on corneal biomechanical properties and
their impact on 24-hour measurement of intraocular pressure. Am. J. Ophthalmol. 2008;
146: 567-572.
[108] Kamiya K, Hagishima M, Fujimura F, Shimizu K. Factors affecting corneal hysteresis
in normal eyes. Graefes Arch. Clin. Exp. Ophthalmol. 2008; 246: 1491-4.
[109] Shah S, Laiquzzaman M, Cunliffe I, Mantry S. The use of the Reichert ocular response
analyser to establish the relationship between ocular hysteresis, corneal resistance factor
and central corneal thickness in normal eyes. Cont. Lens. Anterior Eye. 2006; 29: 257-
62.
[110] Montard R, Kopito R, Touzeau O, Allouch C, Letaief I, Borderie V, Laroche L. Ocular
response analyzer: feasibility study and correlation with normal eyes. J. Fr. Ophtalmol.
2007; 30: 978-84.
[111] Fontes BM, Ambrósio R Jr, Alonso RS, Jardim D, Velarde GC, Nosé W. Corneal
biomechanical metrics in eyes with refraction of -19.00 to +9.00 D in healthy Brazilian
patients. J. Refract. Surg. 2008; 24: 941-5.
[112] Kirwan C, O'Keefe M, Lanigan B. Corneal hysteresis and intraocular pressure
measurement in children using the reichert ocular response analyzer. Am. J.
Ophthalmol. 2006; 142: 990-2.
[113] Lim L, Gazzard G, Chan YH, Fong A, Kotecha A, Sim EL, Tan D, Tong L, Saw SM.
Cornea biomechanical characteristics and their correlates with refractive error in
Singaporean children. Invest. Ophthalmol. Vis. Sci. 2008; 49: 3852-7.
[114] Touboul D, Roberts C, Kérautret J, Garra C, Maurice-Tison S, Saubusse E, Colin J.
Correlations between corneal hysteresis, intraocular pressure, and corneal central
pachymetry. J. Cataract. Refract. Surg. 2008; 34: 616-22.
[115] Ang GS, Bochmann F, Townend J, Azuara-Blanco A. Corneal biomechanical
properties in primary open angle glaucoma and normal tension glaucoma. J. Glaucoma.
2008; 17: 259-62.
[116] Schroeder B, Hager A, Kutschan A, Wiegand W. .Measurement of viscoelastic corneal
parameters (corneal hysteresis) in patients with primary open angle glaucoma.
Ophthalmologe. 2008; 105: 916-20.
[117] Shah S, Laiquzzaman M, Mantry S, Cunliffe I. Ocular response analyser to assess
hysteresis and corneal resistance factor in low tension, open angle glaucoma and ocular
hypertension. Clin. Experiment. Ophthalmol. 2008; 36: 508-13.
[118] Sullivan-Mee M, Billingsley SC, Patel AD, Halverson KD, Alldredge BR, Qualls C.
Ocular Response Analyzer in subjects with and without glaucoma. Optom. Vis Sci.
2008 ; 85: 463-70.
[119] Mollan SP, Wolffsohn JS, Nessim M, Laiquzzaman M, Sivakumar S, Hartley S, Shah
S. Accuracy of Goldmann, ocular response analyser, Pascal and TonoPen XL
tonometry in keratoconic and normal eyes. Br. J. Ophthalmol. 2008; 92: 1661-5.
[120] Ortiz D, Piñero D, Shabayek MH, Arnalich-Montiel F, Alió JL. Corneal biomechanical
properties in normal, post-laser in situ keratomileusis, and keratoconic eyes. J.
[121] Shah S, Laiquzzaman M. Comparison of corneal biomechanics in pre and post-
refractive surgery and keratoconic eyes by Ocular Response Analyser. Cont Lens
Anterior Eye. (In press).
[122] Jaycock PD, Lobo L, Ibrahim J, Tyrer J, Marshall J. Interferometric technique to
measure biomechanical changes in the cornea induced by refractive surgery. J.
[123] Pepose JS, Feigenbaum SK, Qazi MA, Sanderson JP, Roberts CJ. Changes in corneal
biomechanics and intraocular pressure following LASIK using static, dynamic, and
noncontact tonometry. Am. J. Ophthalmol. 2007; 143: 39-47.
[124] Shah S, Laiquzzaman M, Yeung I, Pan X, Roberts. The use of the Ocular Response
Analyser to determine corneal hysteresis in eyes before and after excimer laser
refractive surgery. Cont. Lens. Anterior Eye(In Press).
[125] Chen MC, Lee N, Bourla N, Hamilton DR. Corneal biomechanical measurements
before and after laser in situ keratomileusis. J. Cataract. Refract. Surg. 2008; 34: 1886-
91.
[126] Pallikaris IG, Kymionis GD, Astyrakakis NI. Corneal ectasia induced by laser in situ
keratomileusis. J. Cataract. Refract. Surg. 2001; 27: 1796-80.
[127] Randleman JB. Post-laser in-situ keratomileusis ectasia: current understanding and
future directions. Curr. Opin. Ophthalmol. 2006; 17: 406-12.
[128] Vinciguerra P, Munoz MI, Camesasca FI, Grizzi F, Roberts C. Long-term follow-up of
ultrathin corneas after surface retreatment with phototherapeutic keratectomy. J.
[129] Baek T, Lee K, Kagaya F, Tomidokoro A, Amano S, Oshika T. Factors affecting the
forward shift of posterior corneal surface after laser in situ keratomileusis.
Ophthalmology. 2001; 108: 317-20.
[130] Ou RJ, Shaw EL, Glasgow BJ. Keratectasia after laser in situ keratomileusis (LASIK):
evaluation of the calculated residual stromal bed thickness. Am. J. Ophthalmol. 2002;
134: 771-3.
[131] Miyata K, Tokunaga T, Nakahara M, Ohtani S, Nejima R, Kiuchi T, Kaji Y, Oshika T.
Residual bed thickness and corneal forward shift after laser in situ keratomileusis. J.
[132] Calkins JL, Hochheimer BF, Stark WJ. Corneal wound healing: holographic stress-test
analysis. Invest. Ophthalmol. Vis. Sci. 1981; 21: 322-34.
[133] Kasprzak H, Förster W, von Bally G. Measurement of elastic modulus of the bovine
cornea by means of holographic interferometry. Part 1. Method and experiment. Optom.
Vis. Sci. 1993; 70: 535-44.
[134] Smolek MK. Holographic interferometry of intact and radially incised human eye-bank
corneas. J. Cataract. Refract. Surg. 1994; 20: 277-86.
Chapter 9
SOME ASPECTS OF THE BIOMECHANICS OF SKILLED

MUSICAL PERFORMANCE
Jessie Chen1 and George Moore2

1
Department of Physiology and Neuroscience,
New York University School of Medicine, New York, NY, USA
2
School of Engineering, University of Southern California, Los Angeles, CA, USA
ABSTRACT
This chapter addresses a fundamental aspect of musical performance in string
players: how the physical ―geography‖ of the instrument and bow, and the
anthropometric dimensions of the player interact to produce the stereotypic motor
behavior.
Several factors determine movement, but, unlike most biomechanical tasks, the
determining outcome here is acoustic. Both upper extremities are involved in the
performance but in very different ways.
The left arm governs the contact position of the fingers on the string, and hence the
pitch of a note. The spatial relations of the instrument to the body, the contact point of the
finger on the string, the length of the fingerboard, and the dimensions of the arm
determine a unique posture, and thus the muscle activation patterns, for the left arm for
each individual player. The left arm has quite different postures in relation to the body
and the instrument for cello players in comparison to violinists.
The bowing (right) arm draws the bow across the string. Its travel velocity is the
principal determinant of loudness, but its distance from the bridge, the contact position
along the bow length, bow pressure, and its angle of attack on the string leave a
noticeable effect on the tone, or timbre. Bowing movements are essentially determined by
the flexion and extension of the elbow, with subtle motions of the shoulder and wrist to
keep the bow moving in a straight line perpendicular to the string. Given a fixed spatial
relation between the body of the performer and the instrument, the posture of the arm for
a given bow/string contact point is uniquely determined.
In addition, the force of gravity plays a role in the control of movement. We show,
however, that it affects cellists and violinists in very different ways. Left arm movements
of cellists are more affected by gravity than those of violinists; whereas gravity affects
the right arm more in violinists.
268 Jessie Chen and George Moore
In this chapter we focus on a series of specialized topics: 1) control of the left arm
during shifting movements; 2) fine control of the left arm during corrections of intonation
errors; 3) coordination between the upper arm and forearm; and 4) coordination between
the left and right arms.
INTRODUCTION
In this chapter we discuss some of the basic biomechanical issues of concern to musicians
who play string instruments. We emphasize here the problems specific to cello, with
additional commentary for violin.
The instruments in the violin family have four strings, tuned to different pitches by
adjusting their tension. Strings on the cello typically have a length of about 70 cm, while on
the violin, the strings are much shorter — about 33 cm. The fundamental frequency of a string
is determined primarily by its freely vibrating length (and its intrinsic physical characteristics
which are not variable), and the pitch to which that string is tuned (Figure 1).
Figure 1. A cellist holds a note using his index finger. The finger placement depresses the string which
makes contact with the copper strip affixed to the fingerboard. The string circuit measures the
resistance of the string between this point of contact and the bridge.
Some Aspects of the Biomechanics of Skilled Musical Performance 269
There are two ways a performer may change from one pitch to another. When the desired
note is a short distance away from the current one, the performer may simply switch to
another finger placed at a location appropriate to the desired pitch. In the finer aspects of
performance, the left (fingering) hand and fingers play a complex and essential role that
especially distinguishes beginners from advanced players. But the hand and fingers do not
account for the larger movements such as those in the second type of movement, and it is
these we shall discuss here. This second type is called a ―shift‖. It is an arm movement that
takes the hand along the fingerboard from one location to another. Such movement can
require a displacement of the hand of many centimeters in association with movements of the
forearm and upper arm, and changes in the posture of the hand. In addition to these
movements along a single string, performers often move across strings.
During shifting movements, the finger releases the string, and the arm rotates to bring the
hand to the new desired location. If the current and the next notes are played using the same
finger on the same string, then that finger essentially moves along a straight line trajectory.
Much of the motion is achieved by flexion or extension of the elbow joint, with appropriate
smaller adjustments of the upper arm (in association with shoulder motion) and the hand. For
most purposes the shift must be executed in less than 200 milliseconds.
At the same time, the bow moves across the string to produce vibrations, thus generating
sound. The velocity of bow movement is the primary determinant of loudness, though other
factors play a role as well. The essential fact is that the bow must move along a line
perpendicular to the string. Typical bow lengths range from 63-66 cm for the violin to a
shorter length of 58-61 cm for the cello. Elbow flexion and extension contribute the largest
amount to bow movement, but important contributions to maintaining bow direction and
string contact are also due to the shoulder, the forearm, and the (right) hand. Bow velocities
may be slow (less than 5 cm/sec) or quite fast (more than 60 cm/sec).
It is important to note that gravity often plays a considerable role in string instrument
performance. The violin bow, for example, may sometimes be oriented almost vertically
when on the E- string. Owing to the slight tilt at which the violin is held, with its left side
elevated, the bow, moving progressively to lower pitch strings (E to A, A to D, D to G)
gradually becomes more horizontal (with respect to gravity). On the G-string, gravity ceases
to play a major role. For the cello, gravity plays a similar, but lesser, role in bowing
movements. In contrast, it contributes a great deal to shifting movements on the cello, far less
on the violin. Thus, gravitational forces generally must be included in the force requirements
of motion of bowing and shifting, and for maintaining arm position when motion ceases.
Despite the presence of relatively large movements of both arms, considerable precision
is also required. In the spatial dimension, fingers of the left hand must make contact with the
string and fingerboard with a precision of millimeters for the resulting note to be on pitch.
The required precision depends on the position of the note along the string: higher pitches
(closer to the bridge) require greater positional precision. For the right arm, the slightest
changes in bow velocity or angle make perceptible changes in the tone of notes. Even a slight
tremor accompanying stage fright is clearly audible. In the temporal dimension, muscles of
different segments in the left arm must be activated in an organized fashion in order to move
the hand from one note to another at a precise time to meet the requirements of the music.
The right upper arm and forearm must also be coordinated in order to keep the bow moving in
a straight trajectory perpendicular to the string at the desired speeds.
Finally, there is also a high demand for coordination between the left and right arms,
because the bowing arm is responsible for the tone, the loudness of the note, and special
acoustic effects, while the left arm is responsible for positioning the fingers on the precise
point on the string for the proper pitch. Whatever the left hand is doing, it is audible if the
bow is moving. Sometimes, when the left arm is shifting, it may not be desirable for the tones
intermediate between the pre-shift and post-shift notes to be audible. The bow must therefore
stop moving between these notes, or greatly reduce its velocity.
Great attention is given by teachers to the ways bowing and shifting movements should
be executed, but the actual movements also depend on the particular dimensions of the
performer. Given the variability of body dimensions between players, movements of arms and
hands will be different — even on the same instrument. Movements easily executed by a
larger player may be difficult for a smaller, and vice versa. Teachers typically pay little
attention to the muscle activation patterns underlying movements, how they should be
employed, and for what purpose. The most general guidance is the encouragement of
relaxation, but this is often a misguided piece of advice. Another precept is naturalness and
optimality, and without question the most skillful performers exhibit this quality. But of what
it consists, and how it should be taught, have always been matters of controversy.
In this chapter we focus on a series of specialized topics: 1) control of the left arm during
shifting movements; 2) fine control of the left arm: intonation and error correction; 3)
coordination between the upper arm and forearm; and 4) coordination between the left and
right arms. Our comments are based, in part, on extensive studies of advanced cello students
at the University of Oregon School of Music. An electrical circuit adapted to a cello allowed
us to track the contact point between the string and the fingerboard resulting from the
placement of the finger on that position. We also used solid-state angular-velocity sensors to
monitor angular motion of the right and/or left forearms and upper arms. Details of the
methodology can be found in our previous publications (Chen et al, 2006; 2008). In addition
we monitored the electrical activity (electromyogram-EMG) of various muscle groups of the
arm (recorded by electrodes attached to the skin overlying each muscle), which we had
observed to be active in controlling the basic movements of interest. Through these
simultaneous measurements we were also able to observe the timing relations between the
upper arm and forearm, as well as the right and left arms during performance.
CONTROL OF THE LEFT ARM DURING SHIFTING MOVEMENTS

Motion during a shifting movement primarily takes place at the elbow joint. On the cello,
movement towards a high note (closer to the bridge) requires an extension at the elbow; while
flexion is involved in movement toward a low note (away from the bridge). On the violin,
where the bridge is closest to the head, the reverse is true. Two primary opposing muscles
control elbow rotation (though many more are also activated). Contraction of the biceps
flexes the elbow; activation of triceps extends the elbow. The resulting accuracy depends on
the precision of the performer‘s sense of the desired pitch and upon the development of the
skill (timing of muscle activation and coordination between the opposing muscles) that
enables the performer, after years of training, to move to the required position.
Figure 2. Finger position and associated muscle activities of a cellist performing alternating B-to-A
shifts. A. Shifts without vibrato. B. Shifts with vibrato. Red trace: finger position. Green trace: EMG
recording of biceps activity. Cyan trace: EMG recording of triceps activity. The arrow indicates a
moment when the finger released the string from contact with the fingerboard.
In Figure 2 we show some of the variables in a relatively simple movement by a cellist

during alternate shifts between a higher note (A=440 Hz) and a lower note (B=246.9 Hz)
along the cello A-string, a straight-line distance of 26.8 cm. The lower note is closer to the
player‘s shoulder than the higher note. The finger contact position is shown in the red trace.
The flat portions of this trace are the periods when the finger stayed in a relatively static
position after the performer had completed the movement. Note B is shown at the top of the
figure; note A closer to the bottom. The activities of biceps (green) and triceps (cyan)
associated with the elbow motion are also shown on the same time scale. To move from A to
B requires, among other adjustments, a flexion of the elbow against the force of gravity. Note
that biceps became active just prior to the shift from A to B, and that the biceps activity
persisted as long as the finger was held at B. This continued activity was necessary for
countering the action of gravity. When the biceps activity diminished, the finger released the
string and the shift to A started (elbow extension).
Triceps activity is even more interesting. For one thing, triceps generated more or less
continuous activity along with biceps during the flexion phase that moved the finger towards
B and during the holding period at B. These nominally antagonist muscles were co-
contracting. When the co-contraction phase ended (prior to the shift to A), triceps fired a
brief, higher-amplitude burst, which initiated elbow extension. Gravity carried the arm the
rest of the way. Towards the end of the shift to A, biceps was activated to stop the motion,
and then helped to maintain the position at A. Note, in this example, that both muscles served
jointly to initiate changes of the elbow angle, to assist the braking of that same action, and
then also to stabilize it at a given position. (The thumb, acting against the neck of the
instrument, can also assist in maintaining a given position).
But these same muscles do even more. In Figure 2B, we show the same performer
shifting between the same two notes. This time, however, the cellist is using vibrato, an
oscillatory movement of the hand and arm that changes the finger contact position slightly,
thereby modulating the basic pitch of the note. The same two muscles are shown again, but
now their activation pattern is quite different. It is clear that when the contact position is at A,
biceps fires a string of impulses in synchrony with the change in contact position. That is,
biceps, in addition to its previous functions, is impulsively changing the elbow angle and,
consequently, finger contact position. Given the huge increase in biceps activity when the
finger is at B, we cannot be certain of its periodic activity in relation to the vibrato. The same
is true for triceps, though there appear to be bursts of impulses in triceps out of phase with
biceps, an indication of the opposing function of these two muscles in vibrato production in
cellists. (Violinists also use vibrato, but often the vibrato is produced primarily using wrist
flexion and extension.)
In addition to the different firing patterns of the two muscles, we also note that the same
muscle is activated to a different degree at each note position. This is more apparent in biceps
(Figure 2B). When producing approximately the same size of oscillation, biceps fired more
intensely when the finger was on the lower note, B (where the biceps was activated at a
shorter length). This is due to the so-called ―length-tension relation‖ of the muscle, which
requires greater or lesser neural activation at different muscle lengths.
FINE CONTROL OF THE LEFT ARM: INTONATION AND ERROR

CORRECTION
In the course of ordinary musical performance, mistakes occur. Non-fixed-pitch
instruments, such as string instruments, are especially challenging for pitch accuracy due to
the lack of visual and tactile cues of note position. Carl Flesch (1923), the famous violinist
and teacher, once claimed that
…so-called ―playing in tune‖ is nothing else but an extremely rapid and skillful
correction of the pitch, which was not accurately ―hit‖ the first time.
Pitch errors will often be corrected immediately after the mistake is perceived. Figures
3A and 3B, illustrate two distinctively different error-correction styles. The performer, a
cellist, was asked to shift alternately between two notes on the A string, B (246.9 Hz) and D
(293.7 Hz), without vibrato, a movement distance of 26.8 cm. Here we see the final approach
to D, and the initial part of the return to B. In Figure 3A is shown a continuous positional
adjustment toward the target pitch.
Figure 3. Three short excerpts showing the contact position of the first finger as one subject shifted to
note D (293.7 Hz) from note B (246.9 Hz), a total distance of 9.7 cm. A. The end of a shift showing a
preponderance of slides toward the final position, which was actually sharp, although the pitch might be
acceptable to the performer. B. The end of a shift where the finger was 2 cm short of the intended point.
A series of discrete steps brought the finger closer to the destination. C. A shift from B to D without
using the bow. The entire movement was quite different, and the initial landing point was scarcely
modified, though it would be ―sharp‖.
The performer initially landed flat, about 1.5 cm from the true D position, and then slid
the finger toward the intended pitch. At 700 ms the finger stopped moving, and this contact
position (though about 0.5 cm sharper than the true D position) was held for about 200 ms
before the finger began to return to B. This type of continuous change of contact position
made up about half the cases where error corrections were observed. This example also
reflects our general finding that a performer‘s target pitch (intonation) in the course of a
single two-minute trial was variable. Pitch errors on the violin require similar corrections but
over considerably smaller distances.
In contrast to the continuous correction, Figure 3B shows that after the initial landing,
about 2 cm from the intended note, there was a delay of about 125 msec and a rapid jump in
contact position about 1 cm closer to the final point. After a similar delay there was another
jump closer to D and this contact position was held for about 200 msec, after which there was
a slide to final point. This was the highest pitch within the note, and was held for only 100
msec or so before the required shift returned the finger to note B. Each example in this figure
lasted but a second.
The pattern shown in Figure 3C is quite different from the other two. Here, the contact
position was held without significant adjustment after the initial landing, even when the
landing position was quite far from the correct note position (> 0.5 cm). The only difference
between this panel and the other two is that the performer in this trial was not allowed to use
the bow: thus no acoustic feedback was available.
In fact, when we asked a group of skilled cellists to shift alternately between note B and
note D without using vibrato, at the rate of about one note per second for two minutes, with
and without using the bow, we found that all of them had far fewer positional adjustments in
the no-bow condition (Chen et al, 2008). This implies that it is the perceived pitch error
provided by acoustic feedback that triggers positional adjustments. Without acoustic feedback
performers could reach only the approximate position of a note, and the landing positions
were far more variable between successive notes.
COORDINATION BETWEEN UPPER ARM AND FOREARM

The basic mechanics of shifting movements can be seen using a simplified geometric
diagram as in Figure 4. The two players in the lower panel shift along the fingerboard and
make a series of identical contact positions (i.e., the same notes). The only difference between
them is the proportions of forearm and upper arm lengths. If we assume that the basic spatial
relations between the performer and the instrument remain constant, and that the finger shifts
on a single string, i.e., along a straight line, then the shift will be produced by coordinated
movements of the upper arm and forearm.
Figure 4. Stick figures of simulated arm postures of players with different upper arm and forearm
lengths. The player on the left has a shorter upper arm and longer forearm than the player on the right,
though their total arm lengths are the same. Top. Bowing movements. Bottom. Shifting movements.
Arm sequences are coded from cold to warm colors for postures from low notes to high notes.
The player on the left has a shorter upper arm and longer forearm than the other player on
the right (the total arm lengths are the same). Clearly, arm postures are quite different when
the players move from the low pitch notes to high pitch notes, though the notes are the same.
For the player on the right, the elbow joint first flexes (the upper arm moves backwards in the
saggital plane) on successive notes and then extends for the later notes. For the player on the
left, all joint angles change monotonically.
It is important to point out that because the elbow joint only permits flexion and
extension, the upper arm and forearm always lie in a plane. The spatial orientation of that
plane may change with rotations at the shoulder joint. The arm postures shown in Figure 4,
lower panel, suggest that the plane in which the arm motion lies is almost parallel to the
player‘s sagittal plane. But with inward medial rotation of the shoulder joint, the arm could lie
in the plane of the fingerboard. Our view of the arm would change but not the fundamental
relationship between the arm and the string. The shoulder joint rotation would take place
along the axis of the string. Indeed, as the player attempts to move on the strings toward his
right side (D, G and C), the arm naturally elevates. But the planar position of the upper arm
and forearm would remain relatively constant for equal distances along the string (as shown
in Figure 4, lower panel). Shifts on the violin are less complex, primarily involving forearm
flexion, but the basic posture of the arm in relation to the instrument is more complex.
Let us now consider the task of producing vibrato at each note position. Vibrato is
produced by making small oscillating changes in the finger contact point over a total distance
of just a few millimeters. The finger never leaves the fingerboard. There are two different
ways of moving the contact point. On the cello, one contribution to the contact position
change involves alternating elbow flexions and extensions, a few degrees at most. A second
contribution can come from oscillating rotations of the forearm (alternating pronations and
supinations), again with the contact position being changed by the rotation without the finger
leaving the fingerboard. Most cellists employ a combination of these two.
But the choice of vibrato motions is constrained by the geometry of the arm and the
instrument. In Figure 4C, a low note on the cello fingerboard (closer to the shoulder), elbow
flexions and extensions would move the finger perpendicular to the string and be ineffective
for vibrato purposes. On the other hand, forearm rotations would move the finger along the
axis of the string. For some players, rotations are the only way to produce vibrato at these
locations. Conversely, at the further reaches along the fingerboard, rotations would move
perpendicular to the string and only flexions and extensions would support vibrato. At
intermediate positions either method would suffice depending on the preferences of the
player, though biomechanical factors may play an important role in the player‘s choice.
In Figure 5, we show activity in the bow arm of a cellist illustrating a special style of
bowing — Martelè — that is impulsive in character. The cellist was producing a series of
notes at the same pitch; the bow direction changed with each note. Motion involving the bow
drawn to the right of the body, away from the instrument, is a so-called ―down-bow‖. Motion
toward the fingerboard is called an ―up-bow‖. In both cases the bow motion was essentially
perpendicular to the string. The top trace (blue) shows the alternating angular velocity of the
right forearm. Negative velocity impulses are down-bows; upward impulses are up-bows. The
bottom trace (green) shows the basic bow contact position with the string: rapid movement
brought the bow to a new position, where it stopped.
Figure 5. Activity in the bow arm and corresponding acoustic signal during performance of a special
style of bowing, ―Martelè‖, that is impulsive in character. Top. Angular velocity of the right elbow.
Negative impulses are down-bows; upward impulses are up-bows. Scale: +/- 150 deg/sec. Middle.
Acoustic signal sampled at 360/sec. Bottom. Estimated elbow angle. Scale: +/- 5 degrees relative to
starting angle.
Regarding the acoustic signal (yellow), each note began with a high amplitude and, after
that brief burst, decayed to silence (depending on the physical properties of the instrument).
That acoustic pattern is the desired effect of this bowing style, where each note is ―attacked‖.
We also note that the angular velocity patterns of the bow strokes were not anti-
symmetric, which is to say that an up-bow motion was not the reverse of the down-bow. What
other subtle differences in overall bow control permitted the similarity in the impulsive
acoustic responses is a matter of some complexity.
This style of cello bowing — rapid movement from one position of the bow to another,
followed by a relatively stationary position — can be analyzed using a simplified geometric
representation of the player, the instrument, and the bow. If we assume that the spatial
relation between the player and the instrument remains constant, and that the bow moves
perpendicular to the string, then Figure 4, top panel, gives a suitable representation. Here we
show the same players as those in the lower panel using the same instrument and bow, and the
same string contact position. The differences in posture are evident and give rise to rather
different overall movements as the notes are played (Figures 4A and 4B). The point here is
that the actual bow motion and the resulting acoustic pattern are the same, but the movement
by which the performer realizes the notes is different.
It is important to note, once again, that the upper arm and forearm define a plane. In
Figure 4, top panel, the plane appears to lie along the plane of the fingerboard. But if the
player moves from one string to another it is necessary to raise or lower the elbow to a
slightly different spatial plane. Nevertheless the biomechanical constraints remain the same:
in changing strings the bow and arm rotate around the axis of the string/fingerboard.
We can also see how the cellist might perform a special modification of this bow motion,
namely a sequence of rapid low-amplitude oscillations about a fixed string contact point: a
bowing style called ―tremolo‖. Normally, this movement is executed by alternating elbow
flexions and extensions. But clearly the required movement cannot be performed when the
hand is closer to the string, for here flexion-extension movements would not keep the bow
perpendicular to the string. The anatomical conditions for the player in Figure 4A make it
relatively simple for tremolo to be produced by the three hand positions furthest from the
string. But for the player in Figure 4B tremolo would only be possible for the furthest position
shown. It is clear that tremolo can best be performed when the bow is near the tip because this
arm posture reduces the degrees of freedom the player has to control down to one joint — the
elbow — while satisfying the requirement of the bow being perpendicular to the string.
COORDINATION BETWEEN RIGHT AND LEFT ARMS

When playing a string instrument, it is essential to have precise coordination between the
right and left arms. In Figure 6, we show a cellist shifting alternately between note B and note
D. In Figure 6A, the blue trace represents the angular velocity of the right (bowing) forearm.
When the trace crosses the axis, the bow was stopped near the tip (following a down-bow) or
near the frog (following an up-bow). The red trace shows the contact position of the finger on
the string as the player shifted quickly and alternately between note B (on the A string
furthest from the bridge) and D, a short distance away (9.7 cm closer to the bridge). The zero-
crossings of bow velocity coincided with the mid-position between notes. The maximum bow
velocities occurred while the new note position was held. The green trace is a measure of bow
position, obtained by integrating the velocity signal. The peaks of bow position corresponded
to the moments of transition between notes.
In Figure 6B, we plot the instantaneous value of finger contact position (left) versus
instantaneous value of right elbow angular velocity. The relationship is straightforward; the
loop runs clockwise. In Figure 6C, we plot, for comparison, the linear velocities of the wrist
(horizontal axis) and elbow of the bowing (right) arm during shifts. Their relationships are far
more complex, and bowing motions often involve complex motions of the wrist such as ulnar
flexion and extension that occur several times in each cycle. The lower half of the basic
―figure-8‖ loop runs clockwise, the upper half counterclockwise.
Figure 7 shows a similar situation in which a player shifted alternately between two
notes, B and A (26.8 cm in distant), on the A string. In red is shown the finger contact
position on the string. For reference, the true positions of notes B and A are also shown. From
this, it can be seen that both notes were slightly too close to the bridge (i.e., would sound
sharp to a perceptive listener). In blue is shown the left forearm angular velocity, part of the
motion moving the finger between the two notes. The integral of the right forearm velocity is
shown in green, an indicator of bow position. The bow motion was roughly sinusoidal.
Changes in bow direction occurred at the peaks (zero angular velocity of the right elbow).
The yellow trace shows the acoustic signal recorded simultaneously.
Figure 6. Right and left arm motions during a shifting trial. A. Bow and fingering movements during
shifts between notes B and D. One bow per note. Red, contact position of the finger on the A string.
Top peaks, note B; lower peaks, note D. Scale: +/- 5 cm from average contact position. Blue, right
forearm angular velocity. Scale: +/- 100 deg/sec. Green, estimated right forearm angle. Scale: +/- 1 deg
from the average angle. Positive impulses are down-bow (elbow extension). B. Loop diagram for left
arm finger contact position (horizontal axis) versus right forearm angular velocity (data from A).
Vertical scale: +/- 100 deg/sec. Horizontal scale: +/- 2.5 cm. C. A loop diagram for the right wrist
velocity (horizontal axis) versus right elbow velocity when a cellist shifted between notes B and A.
Horizontal scale: +/- 100 cm/sec. Vertical scale: +/- 100 cm/sec. Positive values correspond to
movements lateral to the mean position of the arm in the horizontal plane. The lower half of the basic
―figure-8‖ loop runs clockwise, the upper half counterclockwise. Data derived from Peak Motion
Analysis System (Peak Performance, Inc., Centennial, CO).
The acoustic traces for each note were quite distinct owing to the fact that they were
aliased by the sampling rate. It can be seen that the acoustic volume was quite attenuated
during the bow changes (arrows), a moment corresponding to a rapid shift in finger position.
The diminished bow velocity at the peaks makes the shift inaudible.
Figure 7. A cellist shifts between notes B and A on the A string. Shift distance: 26.8 cm. Red, finger
contact position. Blue, left forearm angular velocity (flexion: negative). Scale: +/- 100 deg/sec. Green,
estimated right forearm angle, an indicator of bow position. Scale: +/- 3 degrees. Yellow, acoustic
signal (in arbitrary units), sampled at 360/sec.
CONCLUSION
Music pedagogy is not a science. Neither teachers, students or advanced performers can
be consciously aware of the details of the extraordinary complexity of movement that is the
basis for skilled musical performance.
Here, we have attempted to show just a few examples of the details of movement
complexity and coordination. It is unlikely that complete physical descriptions of
performance are even possible. If a musician were to repeat a performance, the details would
certainly not be the same, because, among other skills, musicians can invent new movements,
new interpretations, and new performances.
Mindful of this, the experienced teacher will encourage careful listening, observation,
practice, and ever-deeper insights into the music. The teacher should only hope and believe
that, somehow, with guidance the gifted student will find a way to produce the desired sound,
even without knowing exactly how it is done.
But teachers also need to be aware that the student must also master a unique relationship
to an instrument that is dictated by the anthropometric constraints of the body, the physical
dimensions of the instrument, the physics of the instrument and the biophysics of the body;
always, however, subject to certain as-yet poorly understood principles of the nervous system.
Biomechanics of musical performance is based on all these.
ACKNOWLEDGMENTS
The authors wish to thank Gary Felsing (Motus Bioengineering, CA, USA) for support
and assistance in the experimental trials. We also wish to thank Marjorie Woollacott and
Steve Pologe for contributions to the studies conducted in University of Oregon.
REFERENCES
Chen, J., Woollacott, M., and Pologe, S. (2006). Accuracy and underlying mechanisms of
shifting movements in cellists. Exp. Brain Res, 174, 467-476.
Chen, J., Woollacott, M., Pologe, S., and Moore, GP. (2008). Pitch and space maps of skilled
cellists: Accuracy, variability and error correction. Exp. Brain Res, 188, 493-503.
Flesch, C. Die Kunst des Violinspiels. Band I. Reis und Erler, Berlin. 1923.
Chapter 10
CONTACT HIP STRESS MEASUREMENTS IN

ORTHOPAEDIC CLINICAL PRACTICE
Blaž Mavčič1, Matej Daniel2, Vane Antolič3, Aleš Iglič4 and

Veronika Kralj-Iglič5
1
Department of Orthopaedic Surgery, University Medical Centre Ljubljana, Slovenia
2
Laboratory of Biomechanics, Faculty of Mechanical Engineering,
Czech Technical University Prague, Czech Republic
3
Department of Orthopaedic Surgery, University Medical Centre Ljubljana, Slovenia
4
Laboratory of Physics, Faculty of Electrical Engineering,
University of Ljubljana, Slovenia
5
Laboratory of Clinical biophysics, Faculty of Medicine,
University of Ljubljana, Slovenia
ABSTRACT
There exist several invasive and noninvasive methods to measure the contact hip
stress but due to their complexity only few have so far been tested in clinical trials with
large numbers of participating subjects. Consequently, the use of contact hip stress
measurements in orthopaedic clinical practice is still in its experimental phase.
Biomechanical studies of human hips based on the analysis of 2-D pelvic
radiographs have turned out to be a reasonable compromise between the measurement
accuracy and the feasibility in clinical setting. Clinical studies have shown significantly
higher values of hip stress in adult dysplastic hips when compared to normal hips. It has
been found that the cumulative hip stress independently predicts the WOMAC score after
29-years of follow up in dysplastic hips and does so better than morphological
radiographic parameters of hip dysplasia or the resultant hip force alone. The
preoperative value of the contact hip stress and the magnitude of its operative correction
have been found predictors of the long term success of the Bernese periacetabular
osteotomy. Elevated shear stress in femoral neck, but not elevated hip contact stress, has
been found to be a risk factor for slipping of the capital femoral epiphysis. A statistically
significant correlation between the contact hip stress and the age at the total arthroplasty
has been shown in a group of hips with idiopathic hip osteoarthritis.
282 Blaž Mavčič, Matej Daniel, Vane Antolič et al.
Through advances in 3-D imaging with MRI and CAT scan, visualisation of the
femoral head coverage and pelvic muscle attachment points has improved considerably.
However, the need to supplement the morphological hip status with biomechanical
analysis remains. The current trend is to combine the kinetic gait measurements of the
resultant hip force with 3-D imaging of the hip weight-bearing surface in order to better
estimate the contact hip stress for a given activity/body position. The added value of such
measurements over 2-D pelvic radiograph analysis has not been established yet in clinical
trials.
INTRODUCTION
Hip stress depends on the magnitude/direction of the resultant hip force, the size of the
weight-bearing surface and the stress distribution across this surface. The interest of
orthopaedic surgeons in possible relationship between the contact stress and the pathological
cartilage conditions has been particularly large for weight-bearing joints with relatively
simple kinematics such as the hip joint. A review of the contact hip stress measurement
methods has been published by Brand et al. [1]. While many methods have been developed
only the noninvasive methods could be used in clinical practice. The aim of this paper is to
review the current trends in the clinical use of methods of contact hip stress assessment.
NONINVASIVE DETERMINATION OF THE RESULTANT HIP FORCE

The resultant hip force can be estimated in static or dynamic conditions. Static
biomechanical models estimate the resultant hip force for a given body position by solving
the static equations for the equilibria of forces and torques [2-4]. The muscles in the static
biomechanical models are assumed to be force generators with fixed coordinates.
Interindividual variability of muscle attachment points can be achieved by linear scaling of
the pelvic configuration in plain anterior-posterior radiographs. In this setting, the standing
anterior-posterior pelvic radiographs are assumed to represent the body position of one-
legged stance. One of the static biomechanical models applied to a large number of subjects
in clinical studies is part of the HIPSTRESS method [4-7]. In this model the one-legged
stance is considered to be the representative position for slow gait as the most frequent
activity in everyday life [8]. In the one-legged stance, abductor activity is needed to maintain
balance on the load bearing leg. The following radiographic parameters are measured on the
anterior-posterior radiographs manually [7]: the interhip distance, the pelvic height, the pelvic
width, the coordinates of the insertion point of abductors on the greater trochanter (Figure 1).
The three-dimensional reference coordinates of the muscle attachment points are taken from a
prototype specimen and they are adjusted by linear scaling with regard to the radiographic
pelvic parameters for each individual hip. The solution of the vector equations for the
equilibria of forces and torques yields the three components of the resultant hip force and the
tensions in the abductor muscles.
Noninvasive estimation of the resultant hip force during dynamic activities in different
body positions requires the use of the dynamic biomechanical models.
Contact Hip Stress Measurements in Orthopaedic Clinical Practice 283
Figure 1. In the HIPSTRESS method of the resultant hip force computation, the following radiographic
parameters are measured from the anterior-posterior radiographs manually: the interhip distance l, the
pelvic height H, the pelvic width C, the coordinates of the insertion point of abductors on the greater
trochanter (x, z). The average of acetabular and femoral head radius r and the Wiberg lateral center-
edge angle CE are used subsequently to estimate the weight-bearing surface and the contact hip stress
distribution.
They are based on laboratory measurements with contact force plates, kinematic data of
body segment motion and subsequent inverse dynamic analysis of the moving segments [9,
10]. Such approach necessarily includes complex muscle models for several body segments
and an appropriate optimization technique to solve the model equations with reduntant muscle
forces.
An example of a dynamic biomechanical model based on gait analysis has been
developed for use in pre-clinical testing [10]. A computer model of the bones and muscles of
the human lower extremities (CT-data, Visible Human, NLM, USA) was scaled to match the
anatomies of four THR patients with telemeterized femoral components. Gait analysis data
for walking and stair climbing were determined simultaneously with in vivo hip contact
forces.
The gait data and the individual musculoskeletal models were then used to calculate the
intersegmental resultant joint forces at the ankle, hip and knee joint, as well as the muscle and
joint contact forces throughout each gait cycle. The calculated hip joint contact forces were
finally validated against the contact forces measured in vivo for both activities.
The latest trend in the noninvasive hip force estimation of individual patients is the
polyamid reversed engineering model based on a computed tomography dataset [11]. Pelvis
and femur of an individual patient are reproduced in polyamid by selective laser sintering.
Hip joint forces can be measured using an experimental setup in which an industrial robot is
exerting hip joint forces and moments representing one-legged stance. Hip extensor and
abductor actuator forces are measured which counterbalanced the joint moments. The
resulting bony model is geometrically accurate, but it does not take into account the joint
incongruencies due to the neglected cartilaginous structures in the model [11].
NONINVASIVE DETERMINATION OF THE HIP WEIGHT-BEARING

SURFACE
In the past when three-dimensional imaging was not yet widely available in the clinical
setting, the femoral head radius and the anterior and/or lateral coverage were the main
parameters to be used in the biomechanical models for the estimation of the weight-bearing
surface. The basic method used for the noninvasive determination of the hip weight-bearing
has been plain radiography in the anterior-posterior projection and the false profile view.
Many radiographic indices have been described in order to evaluate the femoral head
coverage by the acetabulum and to indirectly estimate the weight-bearing surface. The most
commonly used parameters include the Wiberg lateral centre-edge angle (CE) [12], the
vertical anterior centre-edge angle of Lequesne and de Sèze (VCA) [13], the femoral head
extrusion index and the acetabular index [14]. Thus, it was assumed femoral head is a perfect
spherical surface with variable lateral and/or anterior coverage by the acetabulum. Soon it
became clear that in most physiological conditions the weight-bearing surface is not equal in
size to the entire articular surface. Variability in some parts of the joint (e.g. lateral coverage)
may greatly influence the hip loading while the medial part of the joint bears only small loads
[6]. In the HIPSTRESS method [4-7], the weight-bearing surface is assumed to make part of a
perfect articular sphere limited on the lateral side by the coverage of the acetabulum. Its
medial border depends on the location of the pole of stress distribution. Thus, the size of the
weight bearing surface is computed from the average of acetabular and femoral head radius
and the Wiberg lateral center-edge angle (Figure 1). The biomechanical role of the horseshoe
geometry of the acetabular cartilage has also described using a three-dimensional
mathematical model. It has been shown the characteristic horseshoe shape of the articular
cartilage in the human acetabulum optimizes the contact stress distribution in the hip joint
[15].
Recent advances in the three-dimensional imaging technologies (CAT scan, MRI) have
greatly improved the ability to estimate the hip articular surface [16]. There exist contact hip
stress studies where the weight-bearing surface was determined with a very precise MRI
imaging albeit the resultant hip force magnitude/direction was assumed to be constant for all
82 hips examined [17]. MRI is particularly valuable for the non-invasive estimation of the
articular cartilage and the labrum. The disadvantage of classic MRI imaging is that (unlike
plain radiography) it cannot be performed with the patient standing and thus the cartilage
deformation in different weight-bearing positions cannot be estimated directly. Although the
three-dimensional image of the articular surface is much more precise it does not suffice to
estimate the weight-bearing surface directly and again biomechanical models must be used to
estimate the importance of different parts of the joint in hip loading. With the development of
MRI and hip arthroscopy, some authors have suggested that impingement due to local
irregularities near the weight-bearing surface may be more important than the contact hip
stress on the weight-bearing surface itself. This has lead to the theory of femoroacetabular
impingement, emphasizing the importance of joint incongruencies and the need to identify
regions of the femoral head surface with locally increased hip stress values due to the
cam/pincer type of impingement [18].
ESTIMATION OF THE CONTACT HIP STRESS IN THE CLINICAL

SETTING
Theoretically, any noninvasive method of the resultant hip force determination could be
combined with any noninvasive method of the hip weight-bearing surface determination
(Figure 2). In the initial research phases the studies mostly focused on the relationship
between the biomechanical parameters based on the analysis of individual cases [1]. In such
setting it was possible to apply very precise methods, although costly in terms of time and
financial resources. However, when the methods were to be applied in clinical studies with
larger numbers of patients (over 100) the aspect of timely/financial feasibility turned out to be
one of the critical factors. Thus, there exist only few studies of contact hip stress performed in
the clinical setting.
Figure 2. The general algorithm of contact hip stress computation includes the determination of the
resultant hip force (combining measurements with the body segment model) and assessment of the
weight-bearing surface with morphological imaging of the joint geometry.
As Brand et al. [1] pointed out reports on contact hip stress in the literature may describe
different stress distributions. Some authors have estimated spatially averaged values of the
contact hip stress (average stress = net force / total weight-bearing surface). Other authors
report values of the peak contact stress, i.e. the maximal contact stress value on the weight-
bearing surface. Several mathematical approaches to estimate the weight-bearing surface (and
consecutively the contact stress distribution) have been proposed [1, 7]. Legal developed a
practical method for calculation of the contact hip joint stress for a specific case based on the
frontal plane equilibrium force analysis [3]. According to this method, the resultant hip joint
force is calculated in static one-legged stance body position assuming one effective abductor
muscle with effective attachment point on the greater trochanter and certain inclination
towards the horizontal plane. Hadley et al. [19] followed the approach outlined by Legal
where the calculation of the hip contact joint stress distribution is restricted to the simplest
case of uniform contact stress distribution. One of the few methods applied to more than two
hundred patients with different types of hip pathology is the HIPSTRESS method [4-7],
developed by the authors of this chapter (Figure 3). The major advantage of the HIPSTRESS
method for evaluation of the stress distribution in the hip joint is that it takes into account the
non-uniform stress distribution over the weight-bearing surface. This could be of special
importance as the gradient of contact stress distribution seems to be even more important than
the magnitude of the contact stress [1]. In the HIPSTRESS model the weight-bearing surface
is not fixed in advance. The hip geometry affects the resultant hip force and the size and the
shape of the weight-bearing surface in a self-consistent manner [6]. These theoretical
predictions are based on the assumption of Hooke's law, where the cartilage is described
macroscopically as a homogeneous continuum and linear elastic solid. This means that the
radial stress in the articular surface is taken to be proportional to the radial strain in the
cartilage layer [2].
In most biomechanical models the femoral head is considered to be a perfect sphere. This
condition may correspond well to the hips with spherical femoral heads, but dysplastic hips in
advanced stages of osteoarthritis have incongruent femoral heads with reduced joint space
width, which leads to radiographic overestimation of the femoral head radius and the Wiberg
lateral center-edge angle and therefore results in underestimation of the peak contact stress
[7]. It is therefore reasonable to expect that incorporation of joint congruity assessment in the
biomechanical model would result in even higher values of peak contact hip stress in
dysplastic hips and therefore improve the predictive value of these biomechanical parameters.
Further, conclusions of biomechanical analyses may be misleading if the relative
mathematical importance of individual biomechanical parameters for the contact hip stress
computation is assumed to be equally important in explaining contact hip stress variability of
the general population. The squared value of the femoral head radius, for example, was found
to have direct inverse correlation to the contact hip stress; yet a clinical study proved the
femoral head radius had small variability between different individuals and did not account
for large contact hip stress differences between normal and dysplastic hips [7].
When critically evaluating the results we should also consider that the statistical
significance of biomechanical computations from anterior-posterior radiographs is limited by
the data dispersion caused by the error in magnification. It was found that the magnification
may vary substantially, but the distribution of the magnification is normal [20].
Figure 3. The HIPSTRESS method is based on the general algorithm of contact hip stress computation,
as shown in Figure 2: the resultant hip force is computed with a static muscle model of one-legged
stance and the joint geometry is assessed by measuring the femoral head radius r and the Wiberg lateral
center-edge angle CE.
Therefore in population studies the magnification should not affect the relative difference
between the average or median values but rather increase the data dispersion and decrease the
statistical significance of the difference between the considered populations. Also, in
determining peak stress, not all the geometrical parameters are equally important as the
functional relations between them are nonlinear. It ensues from the mathematical model that
stress first depends on the radius of the femoral head, then on the interhip distance,
subsequently on the lateral extension of the effective attachment point on the greater
trochanter and at last on the pelvic height and width. For example, if the error made in
determination of the pelvic height is about 15 per cent, this only yields a 2 per cent error in
the peak stress [21].
Three-dimensional imaging techniques have further improved the assessment of the
contact hip stress distribution. Hip morphology data can be applied to the finite element
models in order to simulate the hip contact stress distribution [22]. Patient-specific, non-
linear, contact finite element models of the hip, constructed from computed tomography
arthrograms using a custom-built meshing program, were subjected to normal gait cycle
loads.
There were significant differences found between the normal control and the
asymptomatic hips, and a trend towards significance between the asymptomatic and
symptomatic hips of patients afflicted with developmental dysplasia of the hip. The
magnitudes of peak cumulative contact pressure differed between apposed articular surfaces.
Bone irregularities caused localized pressure elevations and an upward trend between chronic
over-pressure exposure and increasing severity of hip dysplasia [22]. However, the method
has not been applied to larger numbers of patients in the clinical studies.
CONTACT HIP STRESS IN THE ASSESSMENT OF DEVELOPMENTAL HIP

DYSPLASIA AND HIP OSTEOARTHRITIS
Contact hip stress measurements have turned out to be particularly useful in the
evaluation of the developmental dysplasia of the hip [1, 7, 23]. With advances in knowledge,
it has been established that the developmental dysplasia of the hip (previously called
―congenital hip dislocation‖) is not a uniform clinical entity but rather a broad continuous
spectrum ranging from asymptomatic dysplastic acetabula to dislocated hips. Because
insufficient acetabular coverage implies the usual hip loads are distributed on a smaller
weight-bearing surface compared with normal hips, biomechanical research has focused on
estimation of the contact stress in the hip rather than simply morphologic evaluation [24]. In
hips with more severe hip dysplasia, several epidemiologic cross-sectional surveys suggested
increased incidence of hip osteoarthritis [25, 26]. Furthermore, hip dysplasia was found to be
one of the independent risk factors for hip osteoarthritis in addition to age and body mass
index [27]. A higher incidence of hip osteoarthritis together with higher average values of
contact stress in dysplastic hips have led to the hypothesis that contact hip stress may be one
of the key parameters involved in cartilage degeneration [1, 2]. Direct clinical assessment of
the predictive value of contact hip stress was reported in two clinical studies of patients with
hip dysplasia who were treated with closed reduction and followed up to the average age of
31 years. These authors concluded increased cumulative stress exposure bears higher risk for
an unfavorable clinical outcome or osteonecrosis [19, 28]. Although the most severe cases of
hip dysplasia are clearly associated with early degeneration, reports on patients with
borderline dysplastic hips have been more controversial. In a study with 10-year follow-up of
age-matched patients with residual dysplasia without subluxation and normal hips, the
authors reported no differences in the reduction of the joint space width or in self-reported hip
pain [29]. A recent systematic review found little evidence for a relationship between hip
dysplasia and late hip osteoarthritis discovered in patients older than 50 years of age [26].
However, the authors recognized the relationship for the subsequent risk of osteoarthritis in
persons diagnosed with dysplasia at a young age compared with the subsequent risk of young
patients with osteoarthritis without dysplasia. Some authors have speculated most of the cases
of ―idiopathic‖ hip osteoarthritis in fact arise as a result of subtle abnormalities in the
anatomic structure of the hip that remained unrecognized during childhood and adolescence
and only began to cause clinical symptoms in old age [30].
A clinical study was conducted on the role of the contact hip stress for the development
of osteoarthritis in initially asymptomatic human hips, either dysplastic or normal [23]. In the
study nonoperatively treated nonsubluxated hips with developmental dysplasia without
symptoms at skeletal maturity were identified and compared to adult hips without any
disease. Peak contact hip stress was computed with the HIPSTRESS method [4-7] using
anterior-posterior pelvic radiographs at skeletal maturity. This method enabled computation
of the peak contact hip stress for every individual hip from known values of the body weight,
the femoral head radius, the Wiberg center-edge angle, the magnitude of the resultant hip
force and the inclination of the resultant hip force with respect to the vertical. The cumulative
contact hip stress was determined by multiplying the peak contact hip stress by age at follow-
up. The WOMAC scores [31] and radiographic indices of osteoarthritis at a minimum follow-
up of 20 years were compared. Dysplastic hips had higher mean peak contact hip stress and
higher mean cumulative contact hip stress than normal hips. Mean WOMAC scores and
percentage of asymptomatic hips in the study group at the average age of 51 years were equal
to the control group at the average age of 68 years. After adjusting for gender and age, the
cumulative contact hip stress, Wiberg center-edge angle, body mass index, but not the peak
contact hip stress, independently predicted the final WOMAC score in dysplastic hips.
Cumulative contact hip stress predicted early hip osteoarthritis better than the Wiberg center-
edge angle [23].
The relative maximum hip joint contact stress was found to be higher in healthy women
than in healthy men [32]. As women have a higher incidence of the hip osteoarthritis [33],
such epidemiological results support the hypothesis that the increased contact hip stress can
be one of the risk factors of hip osteoarthritis [32]. This hypothesis was tested by two studies
of standard anterior-posterior pelvic radiographs with no or subtle radiological signs of hip
osteoarthritis of patients, who underwent hip arthroplasty for primary osteoarthritis years later
[34, 35]. In the population of subjects with unilateral osteoarthritis, average peak contact hip
stress was significantly higher in hips with arthroplasty than in contralateral hips. In the
population of subjects with bilateral osteoarthritis, average peak contact hip stress was
significantly higher in hips with earlier arthroplasty than in contralateral hips [34]. Younger
age at hip arthroplasty was associated with higher body weight, higher peak contact hip stress
normalized to the body weight, higher resultant hip force and larger peak contact hip stress,
but not with body mass index [35].
CONTACT HIP STRESS IN PREOPERATIVE PLANNING OF

ORTHOPAEDIC PROCEDURES
Osteoarthritis can develop as an idiopathic disease, however, subtle abnormalities could
be detected in the hip joint prior to the development of symptoms. Origins of the development
of osteoarthritis are ascribed to metabolic resorption of cartilage and/or deformations of
anatomical structures. The deviations in the size, shape, mutual proportions or orientation of
the acetabulum and/or the femoral head occur frequently. Such deviations are described as hip
dysplasia [19]. Although mostly a pediatric problem, hip dysplasia can persist in untreated or
unsuccessfully treated cases into adulthood as residual hip dysplasia and may eventually lead
to degeneration of the cartilage, presumably due to the pathologically increased stress within
the joint. Therefore hip dysplasia represents an important indication for operative procedures
that should reduce or redistribute the hip joint stress, thereby stopping or slowing down the
pathological processes in the hip cartilage. The hypothesis of secondary osteoarthritis due to
decreased femoral head coverage and consecutively higher contact hip stress [28, 36] has led
to the invention of surgical procedures of acetabular reorientation in order to prevent the
progression of osteoarthritis. Several operative procedures to increase femoral head coverage
have been described [37-40] with limited possibilities of correction. Because of such
limitations the ‗Bernese‘ periacetabular osteotomy was developed in 1984 and published in
1988 [41]. The procedure has become widely used because it allows optimal correction with
minimal exposure and low complication rate, although the operation is technically demanding
and its learning curve is long [42]. Ideally this procedure would be indicated in younger
adults with concentric hip motion, spherical joint surfaces and no secondary osteoarthritis, but
many patients who present with symptomatic hip dysplasia do not meet these criteria [43].
Many attempts have been made to use additional clinical, radiographic or biomechanical
factors to carefully select patients that would benefit from the joint preserving surgery [44].
Biomechanical parameters have also been used in the preoperative planning in order to
determine the optimal degree of correction in different planes and to achieve the subtle
balance between improvement of femoral coverage and restriction of range of motion [5, 16,
17].
In patients who underwent the Bernese periacetabular osteotomy there have been many
retrospective analyses published, but only two of them have so far included the pre- and
postoperative computations of the contact hip stress [45, 46]. Both studies used the
HIPSTRESS method applied to 170 hips altogether. The periacetabular osteotomy was shown
to improve the lateral and anterior coverage of the femoral head and accordingly to reduce the
normalized peak contact stress in all studied hips. Biomechanical results were consistent with
previous studies that have shown contact pressure elevation in dysplastic hips [7, 19] and
contact pressure reduction by simulated acetabular reorientation [17]. The postoperative
values were in most cases reduced to the level observed in healthy adult hips [47]. Although
the Bernese periacetabular osteotomy allowed medialization of the hip joint in addition to
improvement of lateral and anterior coverage [48], some studied hips showed lateralization of
the center of rotation on the postoperative AP radiographs. Accordingly, the magnitude of the
resultant hip force increased in some stratification groups, which is consistent with theoretical
predictions of the relationship between the center of rotation and the resultant hip force [5].
These clinical results therefore confirmed theoretical findings [5] that the resultant hip force
itself is not a sufficient biomechanical parameter for preoperative planning and that rotational
osteotomy can effectively decrease the contact hip stress although the resultant hip force may
be slightly increased. The conclusions of this study have shown further reduction of the
contact hip stress is only possible through medialization of the femoral head and not through
excessive lateral coverage. Accordingly, the procedure was modified with curved
periacetabular osteotomy that enables better medialization of the femoral head [46].
Contact hip stress has been tested as a risk factor for slipped capital femoral epiphysis,
but with less success [49]. Many hypotheses about the etiology of slippage have been
examined, yet the underlying mechanisms have not yet been fully elucidated. Hips
contralateral to the slipped ones were compared to age- and gender-matched healthy hips with
respect to the shear stress and the contact hip stress. The characteristics of individual hips
were incorporated by means of geometrical parameters determined from standard anterior-
posterior radiographs. Shear stress was calculated by using a mathematical model where the
femoral neck was considered to function as an elastic rod. Contact hip stress was calculated
by the HIPSTRESS method [4-7]. Hips contralateral to the slipped ones had higher average
shear stress in the femoral neck and more vertically inclined physeal angle in comparison to
healthy hips, shear stress in the contralateral hips to the slipped ones remained significantly
higher even when normalized to the body weight. However, there was no significant
difference in the average contact hip stress [49].
The long-term effect of contact hip stress on the clinical outcome was also studied in the
hips operated on by various intertrochanteric osteotomies due to avascular necrosis of the
femoral head [50]. The hypothesis stated the hips with a more favorable postoperative
distribution of contact hip stress had better clinical outcome. For each hip the peak contact hip
stress before/after the operation was determined with the modification of the HIPSTRESS
method that took into account the non-weight-bearing necrotic part of the femoral head. The
hips were evaluated clinically 9-26 years after the operation and divided into a successful and
an unsuccessful group. In the successful group the operation caused an average decrease of
the peak hip stress of about 10%, while in the unsuccessful group the operation caused an
average increase of the peak hip stress of about 4%, the difference between the respective
changes of the peak stress due to the operation being statistically significant [50].
CONCLUSION
The development of noninvasive models to estimate contact hip stress and clinical
application of these models have been running hand in hand. Biomechanical studies of human
hips based on the analysis of the two-dimensional pelvic radiographs have turned out to be a
reasonable compromise between the measurement accuracy and the feasibility in clinical
setting. Through wider availability of the three-dimensional imaging techniques, visualization
of the femoral head coverage and pelvic muscle attachment points has improved
considerably. However, the need to supplement the morphological hip status with
biomechanical analysis remains. One of the limitations of the present biomechanical contact
hip stress computations is the inability to estimate the patients‘ physical activity levels
throughout their lifetime as such data cannot be reliably acquired retrospectively. The
potential confounding effect of differences in physical activity is difficult to estimate because
there is no clear evidence lifelong standing, walking, or lifting in the population with normal
hips is a risk factor for osteoarthritis. The current trend in biomechanics is to combine the
kinetic gait measurements of the resultant hip force with three-dimensional imaging of the hip
weight-bearing surface in order to better estimate the contact hip stress for a given
activity/body position. Nevertheless, the added value of such measurements over the two-
dimensional pelvic radiograph analysis still needs to be established by clinical trials with
larger numbers of participating subjects. Eventually, the clinical use of the contact hip stress
methods will depend both on their accuracy and feasibility.
REFERENCES
[1] Brand RA, Iglič A, Kralj-Iglič V. Contact stress in the human hip, implications for
disease and treatment. Hip. Int. 2001;11:117-26.
[2] Brinckmann P, Frobin W, Hierholzer E. Stress on the articular surface of the hip joint in
healthy adults and person with idiopathic osteoarthrosis of the hip joint. J. Biomech.
1981;14:149-56.
[3] Legal H: Introduction to the Biomechanics of the Hip. In: Congenital Dysplasia and
Dislocation of the Hip. Ed by Tönnis D. Berlin, Springer-Verlag, 1987: pp 26-57.
[4] Iglič A, Srakar F, Antolič V. Influence of the pelvic shape on the biomechanical status
of the hip. Clin. Biomech. 1993;8:223-4.
[5] Iglič A, Kralj-Iglič V, Antolič V, Srakar F, Stanič U. Effect of the periacetabular
osteotomy on the stress on the human hip joint articular surgace. IEEE Trans. Rehab.
Engr. 1993;1:207-12.
[6] Ipavec M, Brand R A, Pedersen D R, Mavčič B, Kralj-Iglič V, Iglič A. Mathematical

modelling of stress in the hip during gait. J. Biomech. 1999;32:1229-35.
[7] Mavčič B, Pompe B, Antolič V, Daniel M, Iglič A, Kralj-Iglič V. Mathematical
estimation of stress distribution in normal and dysplastic human hips. J. Orthop. Res.
2002;20:1025-30.
[8] McLeish RD, Charnley J. Abduction forces in the one-legged stance. J. Biomech.
1970;3:191-209.
[9] Hashimoto N, Ando M, Yayama T, Uchida K, Kobayashi S, Negoro K, Baba H.
Dynamic analysis of the resultant force acting on the hip joint during level walking.
Artif. Organs. 2005;29:387-92.
[10] Heller MO, Bergmann G, Kassi JP, Claes L, Haas NP, Duda GN. Determination of
muscle loading at the hip joint for use in pre-clinical testing. J. Biomech. 2005;38:1155-
63.
[11] Pfeifer R, Hurschler C, Ostermeier S, Windhagen H, Pressel T. In vitro investigation of
biomechanical changes of the hip after Salter pelvic osteotomy. Clin. Biomech.
2008;23:299-304.
[12] Wiberg G. Studies on dysplastic acetabula and congenital subluxation of the hip joint-
with special reference to the complication of osteoarthritis. Acta. Orthop. Scand.
1939;83(Suppl 58):1-135.
[13] Lequesne M, de Sèze S. Le faux profil du bassin: Nouvelle incidence radiographique
pour l'étude de la hanche: Son utilité dans les dysplasies et les différentes coxopathies.
Rev. Rhum. Mal. Osteoartic. 1961;28:643-52.
[14] Delaunay S, Dussault RG, Kaplan PA, Alford BA. Radiographic measurements of
dysplastic adult hips. Skeletal. Radiol. 1997;26:75-81.
[15] Daniel M, Iglič A, Kralj-Iglič V. The shape of acetabular cartilage optimizes hip
contact stress distribution. J. Anat. 2005;207:85-91.
[16] Mechlenburg I, Nyengaard J R, Romer L, Søballe K. Changes in load-bearing area after
Ganz periacetabular osteotomy evaluated by multislice CT scanning and stereology.
Acta. Orthop. Scand. 2004;75:147-53.
[17] Hipp JA, Sugano N, Millis MB, Murphy SB. Planning acetabular redirection
osteotomies based on joint contact pressures. Clin. Orthop. Relat. Res. 1999;364:134-
43.
[18] Ganz R, Leunig M, Leunig-Ganz K, Harris WH. The etiology of osteoarthritis of the
hip: an integrated mechanical concept. Clin. Orthop. Relat. Res. 2008;466:264-72.
[19] Hadley NA, Brown TD, Weinstein SL. The effects of contact pressure elevations and
aseptic necrosis on the long-term clinical outcome of congenital hip dislocation. J.
Orthop. Res. 1990;8:504-13.
[20] Vengust R, Drobnič M, Daniel M, Antolič V, Pernus F, Iglič A, Kralj-Iglič V. Role of
magnification of standard anterior-posterior radiographs in determination of contact hip
joint stress. Biomed. Eng. Appl. Basis. Comm. 2000;12:242-4.
[21] Daniel M, Antolič V, Iglič A, Kralj-Iglič V. Determination of contact hip stress from
nomograms based on mathematical model. Med. Eng. Phys. 2001;23:347-57.
[22] Russell ME, Shivanna KH, Grosland NM, Pedersen DR. Cartilage contact pressure
elevations in dysplastic hips: a chronic overload model. J. Orthop. Surg. Res. 2006;3:1-
6.
[23] Mavčič B, Iglič A, Kralj-Iglič V, Brand RA, Vengust R. Cumulative hip contact stress
predicts osteoarthritis in DDH. Clin. Orthop. Relat. Res. 2008;466:884-91.
[24] Brand RA, Pedersen DR, Davy DT, Kotzar GM, Heiple KG, Goldberg VM.
Comparison of hip force calculation and measurements in the same patient. J.
Arthroplasty 1994;9:45-51.
[25] Jacobsen S, Sonne-Holm S. Hip dysplasia: a significant risk factor for the development
of hip OA. A cross-sectional survey. Rheumatology 2005;44:211-8.
[26] Lievense AM, Bierma-Zeinstra SMA, Verhagen AP, Verhaar JAN, Koes BW.
Influence of hip dysplasia on the development of OA of the hip. Ann. Rheum. Dis.
2004;63:621-6.
[27] Jacobsen S, Sonne-Holm S. Increased body mass index is a predisposition for treatment
by total hip replacement. Int. Orthop. 2005;29:229-34.
[28] Maxian TA, Brown TD, Weinstein SL. Chronic stress tolerance levels for human
articular cartilage: two non-uniform contact models applied to long term follow up of
CDH. J. Biomech. 1995;28:159-66.
[29] Jacobsen S, Sonne-Holm S, Søballe K, Gebuhr P, Lund B. Joint space width in
dysplasia of the hip. J. Bone Joint Surg. Br. 2005;87:471-7.
[30] Murphy SB, Kijewski PK, Millis MB, Harless A. Acetabular dysplasia in the
adolescent and young adult. Clin. Orthop. Relat. Res. 1990;261:214-23.
[31] Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of
WOMAC: a health status instrument for measuring clinically important patient relevant
outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or
knee. J. Rheumatol. 1988;15:1833-40.
[32] Iglič A, Daniel M, Kralj-Iglič V, Antolič V, Jaklič A. Peak hip joint contact stress in
male and female populations. J. Musculoskeletal. Res. 2001;5:17-21.
[33] Heliovara M, Makela M, Impivaara O, Knekt P, Aromaa A, Sievers K. Association of
everweight, trauma and workload with coxaarthrosis. A health survey of 7217 persons.
Acta. Orthop. Scand. 1993;64:513-8.
[34] Rečnik G, Kralj-Iglič V, Iglič A, Antolič V, Kramberger S, Vengust R. Higher peak
contact hip stress predetermines the side of hip involved in idiopathic osteoarthritis.
Clin. Biomech. 2007;22:1119-24.
[35] Rečnik G, Kralj-Iglič V, Iglič A, Antolič V, Kramberger S, Rigler I, Pompe B, Vengust
R The role of obesity, biomechanical constitution of the pelvis and contact joint stress
in progression of hip osteoarthritis. Osteoarthritis Cartilage 2009 (in print).
[36] Michaeli DA, Murphy SB, Hipp JA. Comparison of predicted and measured contact
pressures in normal and dysplastic hips. Med. Eng. Phys. 1997;19:180-6.
[37] Salter RB. Innominate osteotomy in the treatment of congenital dislocation and
subluxation of the hip. J. Bone Joint Surg. Br. 1961;43-B:518-39.
[38] LeCoeur P. Corrections des défauts d'orientation de l'articulation coxo-fémorale par
ostéotomie de l'isthme iliaque. Rev. Chir. Orthop. 1965;51:211-2.
[39] Chiari K. Medial displacement osteotomy of the pelvis. Clin. Orthop. Relat. Res.
1974;98:55-71.
[40] Tönnis D. Congenital dysplasia and dislocation of the hip in children and adults. New
York, Springer 1987.
[41] Ganz R, Klaue K, Vinh TS, Mast JW. A new periacetabular osteotomy for the treatment
of hip dysplasia. Technique and preliminary results. Clin. Orthop. Relat. Res.
1988;232:26-36.
[42] Søballe K. Pelvic osteotomy for acetabular dysplasia. Acta. Orthop. Scand.
2003;74:117-8.
[43] Murphy S, Deshmukh R. Periacetabular osteotomy: preoperative radiographic
predictors of outcome. Clin. Orthop. Relat. Res. 2002;405:168-74.
[44] Trousdale RT, Cabanela ME. Lessons learned after more than 250 periacetabular
osteotomies. Acta. Orthop. Scand. 2003;74:119-26.
[45] Kralj M, Mavčič B, Antolič V, Iglič A, Kralj-Iglič V. The Bernese periacetabular
osteotomy: clinical, radiographic and mechanical 7-15-year follow-up of 26 hips. Acta.
Orthop. 2005;76:833-40.
[46] Teratani T, Naito M, Shiramizu K, Nakamura Y, Moriyama S. Modified pubic
osteotomy for medialization of the femoral head in periacetabular osteotomy: a
retrospective study of 144 hips. Acta. Orthop. 2008;79:474-82.
[47] Mavčič B, Slivnik T, Antolič V, Iglič A, Kralj-Iglič V. Hip contact stress is related to
the development of hip pathology with increasing age. Clin. Biomech. 2004;19:939-43.
[48] Crockarell J Jr, Trousdale RT, Cabanela ME, Berry DJ. Early experience and results
with the periacetabular osteotomy. The Mayo Clinic experience. Clin. Orthop. Relat.
Res. 1999;363:45-53.
[49] Zupanc O, Krizančič M, Daniel M, Mavčič B, Antolič V, Iglič A, Kralj-Iglič V. Shear
stress in epiphyseal growth plate is a risk factor for slipped capital femoral epiphysis. J.
Pediatr. Orthop. 2008;28:444-51.
[50] Dolinar D, Antolič V, Herman S, Iglič A, Kralj-Iglič V, Pavlovčič V. Influence of
contact hip stress on the outcome of surgical treatment of hips affected by avascular
necrosis. Arch. Orthop. Trauma. Surg. 2003;123:509-13.
Chapter 11
EXTERNAL PELVIC FIXATION DURING LUMBAR

MUSCLE RESISTANCE EXERCISE
Michael C. McGlaughlin1, Philip A. Anloague1 and Brian C. Clark2

1
Physical Therapy Program, Department of Health and Sport Science,
University of Dayton and University of Dayton, Dayton, OH; USA
2
Institute for Neuromusculoskeletal Research, Department of Biomedical Sciences,
Ohio University College of Osteopathic Medicine, Athens, OH, USA
ABSTRACT
Resistance exercise has long been used to promote musculoskeletal health with the
application of training regimens for the clinical treatment and prevention of low back
pain growing in popularity over the last couple of decades. A variety of exercise modes
have been utilized in an attempt to stimulate and promote increases in muscle function of
the lumbar extensors. This chapter examines the current state of knowledge regarding the
application of external pelvic fixation during trunk extension exercise and its importance
on the concomitant increase of functional outcomes such as muscle strength, muscle
activation patterns and compensatory muscle growth.
Keywords: lumbar, exercise, muscle, low back pain.
INTRODUCTION
It has been suggested that lumbar muscle training aids in increasing pelvic and spinal
stability [1], and over the past few decades a plethora of scientific interest has developed
regarding the optimal way to effectively strengthen the integrity of the lumbar musculature
that surrounds the spine. In 1994, the US Agency for Health Care and Policy Research (now

Corresponding Author and Reprint Requests to: Brian C. Clark, Ph.D. Dept Biomedical Sciences Ohio University
211 Irvine Hall, Athens, OH 45701, 740-593-2354, Email: clarkb2@ohio.edu
296 Michael C. McGlaughlin, Philip A. Anloague and Brian C. Clark
the Agency for Healthcare Research and Policy) concluded that back extensor muscle
conditioning exercises were helpful in the management of acute low back pain (LBP) [2].
However, determining the most effective protocol to strengthen the lumbar musculature has
proven somewhat difficult, as researchers have reported wide variations on the effectiveness
of exercise prescriptions in increasing lumbar muscle mass and strength. While much of this
variation is likely due to the between-study differences in the exercise prescription employed
(i.e. training frequency, intensity and global versus targeted training), one potential factor
involved in effectively increasing lumbar extensor muscle function surrounds the
biomechanics of training and the degree to which the muscles are effectively isolated and
targeted during resistance training exercises.
Typically, extension of the trunk is a coupled movement initiated by extension of the hips
followed by extension of the lumbar spine [26]. This normal lumbo-pelvic rhythm reduces the
demands on the lumbar extensor muscles and underlying apophyseal joints and discs, thereby
protecting the region against high stress [27]. Delay in lumbar extension shifts the extensor
torque demand to the powerful hip extensors (hamstrings and gluteus maximus), at the time
when the external flexion torque on the lumbar region is greatest [27]. This phenomenon
creates difficulty in accurately quantifying range-of-motion and torque production making it
difficult to determine the relative contribution of each joint segment on these measures [28].
Pollack, Graves and others have previously suggested that to strengthen the lumbar
muscles external pelvic fixation (EPF) is required whereby the lumbar extensors are activated
in an isolated fashion to limit the compound lumbar-pelvic rhythm and the contributing
effects of the gluteals and hamstrings which have a much larger cross sectional area (CSA)
and larger moment arms (Figure 1) [3-6,9-14].
Reprinted with permission from Carpenter and Nelson [6].
Figure 1. During compound trunk extension the contraction of the gluteal and hamstring muscles can
produce around 110 degrees of pelvic rotation about the femurs, while simultaneous contraction of the
lumbar extensor muscles can produce around 72 degrees of lumbar vertebral movement.
External Pelvic Fixation during Lumbar Muscle Resistance Exercise 297
Reprinted with permission from Clark and colleagues [23].
Figure 2. Electromyogram (EMG) activity of the lumbar and hip extensor muscles during fatiguing
trunk extension exercise without pelvic restriction. Note the plateau and subsequent decline in lumbar
muscle EMG amplitude when the subjects reached approximately 55% of their maximal fatigue
capacity, and the concomitant increase in hip extensor activity. This observation indicates that during
lumbar extension without pelvic restriction the hip extensors accommodate the performance of the task.
A similar observation has also been made with an increase in external loading [22, 25].
For example, compound trunk extension involves both contraction of the hip extensor
muscles (gluteal and hamstrings) that can result in ~ 110 degrees of pelvic rotation about the
femur as well as a contraction of the lumbar extensor musculature that can result in ~ 72
degrees of lumbar vertebral movement (Figure 1) [6].
For the purposes of this chapter, the term external pelvic fixation (EPF) has been adopted
so that the reader will not be confused with various interchangeable terms, such as neutral
pelvis, lumbo-pelvic stabilization, isolated lumbar extension, pelvic restraint, pelvic isolation,
or spinal stabilization which have been used when referring to mechanisms performed with
internal structures or through external methods [1,4,5,10,11,16,20,21,29,30,31,32].
In this chapter we will discuss the effect that external fixation of pelvic motion has during
resistance training on the exercise-induced physiologic responses and muscle performance
adaptations. Specifically, we will review the effect of EPF on the exercise training increases
in neuromuscular strength and muscle mass as well as its effect on muscle activation patterns.
EFFECT OF PELVIC FIXATION ON TRAINING-INDUCED INCREASES IN

STRENGTH
When most skeletal muscles are exposed to short-term (i.e. 8-10 weeks) heavy resistance
exercise training they tend to increase their maximal voluntary force output (strength)
between 20-30% [7, 8]. With regards to the lumbar extensor muscles however, large
variations in reported strength gains exist with some authors reporting no increases in strength
[9] and others observing greater than 100% increases in strength [4, 5, 10, 11, 30]. While
there could be a number of factors explaining these discrepant results one postulated reason is
associated with the degree of EPF that is provided during the resistance training protocol.
Graves and colleagues investigated whether pelvic stabilization was imperative to
strengthening the lumbar extensor muscles and observed that 12-weeks of variable resistance
training with EPF (1 day/wk) resulted in a significant mean increase of 23.5% in isolated
lumbar extension strength (strength testing was conducted with EPF imposed to minimize
synergistic contributions of the hip extensors) [4].
Interestingly, subjects who performed the resistance training protocol without EPF did
not significantly increase their isolated lumbar extension strength despite increasing their
training load leading the authors to suggest that trunk extension exercise without EPF
restriction primarily exercised the gluteal and hamstring muscles as they are heavily involved
in pelvic rotation. Similarly, Mayer et al. did not observe significant increases in strength
when trunk extension training did not include EPF [9]. In this study, subjects performed
Roman chair exercise (a common exercise device found in fitness and rehabilitation settings)
which involved progressive resistance back extension exercise on a device that allows one to
lift and lower their upper body with variable resistance added by using hand-held metal
plates. While subjects performed the training on the Roman chair (which does not limit pelvic
rotation) strength testing was performed on a device that does limit the hip extensor
contribution to maximal force.
The aforementioned studies all utilized healthy individuals without low back pain.
However, a number of authors trained symptomatic patient populations with EPF and found
similar results with respect to strength increases, although to a lesser magnitude in
comparison to that observed in healthy individuals [3, 12-14].
These patient populations were quite varied including geriatric women [13], a post-
surgical population after just 6 weeks recovery from lumbar discectomy [3], and groups with
chronic low back pain [12, 14]. When the results of the aforementioned studies performing
resistance exercise training with EPF are considered collectively, the mean strength increase
is on the order of ~ 50%.
This magnitude is exceptional, especially when one considers that the training frequency
in many of these studies was as low as one or two times per week. Other evidence of
improvements with EPF in the symptomatic population has been documented in the literature,
with one dramatic example by Deutsch reporting increases in lumbar extensor strength greater
than 300% throughout the range-of-motion with an 8 week rehabilitation training protocol
[30].
EFFECT OF PELVIC RESTRICTION ON TRAINING-INDUCED INCREASES

IN MUSCLE MASS
A common outcome variable to determine the effectiveness of a resistance training

protocol on activating and stimulating a target muscle is the assessment of muscle mass (most
commonly cross-sectional area (CSA) or volume). An increased muscle mass is primarily
resultant of an increase in the contractile proteins actin and myosin and is in part responsible
for the increases in strength described above. Relatively few researchers have investigated
CSA changes in the lumbar extensors with EPF exercise training and to our knowledge no
one has directly compared exercise training regimens with and without EPF for this outcome
variable. Li and colleagues reported that 1 set of resistance exercise at 80% of maximal
strength 1 time per week for 12-weeks resulted in a 8% increase in lumbar extensor CSA
while three times per week increased CSA 15% [15]. These morphological increases were
accompanied with concomitant peak torque improvements of 31-43%. While it is difficult to
determine the true effect of EPF on these morphological adaptations without having a
comparison group performing the same protocol without restriction, these observations do
suggest that lumbar muscle growth can be induced with a relative small degree of muscle
overload when training with EPF is performed. More recently, Danneels et al. (2001) reported
CSA changes in the multifidus muscle following 10-weeks of exercise training in chronic low
back pain patients [16]. They found that dynamic core stability exercises such as leg
extension in the four point kneeling position and trunk extension exercise without EPF did
not significantly increase multifidus CSA, but when the same exercises were performed with
a 5-sec isometric contraction separating the concentric and eccentric contraction phases they
observed that multifidus CSA increased ~ 7%. Interestingly, the static component that they
observed to be critical for compensatory muscle growth could in effect be construed as
‗pelvic fixation,‘ albeit not external, since the isometric contractions are indeed non-moving
and perhaps serving to minimize synergistic contribution from accessory muscles.
MUSCLE ACTIVATION PATTERNS DURING TRUNK EXTENSION

EXERCISE
While the efficacy of training paradigms can only directly be determined with a clinical
outcome trials study design, one can glean information on the effectiveness of different
exercise tasks from acute studies evaluating the muscle activation patterns of synergistic
muscles. The most common way to assess muscle activation patterns is through the recording
and evaluation of the electromyogram (EMG) signal which is well-documented to have an
increased amplitude associated with the recruitment of motor units and increases in discharge
rate [17]; although another method to evaluate muscle activity is though the analysis of the
transverse relaxation time (T2) obtained from magnetic resonances images, as a slowing in
the T2 following exercise is associated with the intensity of skeletal muscle activation [18,
19].
As stated above, it appears that EPF is required to increase voluntary force output from
the lumbar extensor muscles [11], thus, one might expect to find greater levels of activation
during lumbar extension exercises with EPF than without. A recent report from San Juan and
colleagues investigated this potential with asymptomatic subjects, and indeed did report that
submaximal lumbar extension exercise (load equal to 50% of maximal strength) with EPF
resulted in a 51% greater amplitude of the surface EMG signal recorded from the lumbar
paraspinals when compared to the EMG amplitude recorded from trunk extension without
EPF (no differences were observed with respect to biceps femoris activity between the two
conditions) [20]. While this finding does suggest that EPF is needed for optimal lumbar
extensor activation, discrepant results do exist. For instance, Udermann and colleagues
reported that pelvic fixation did not influence the muscle activation of the lumbar extensors
during submaximal dynamic contractions [21], however, it is possible that the restraint
condition imposed in this study did result in some pelvic restriction as they did apply restraint
devices, but did not completely tighten them (as opposed to the San Juan paper which did not
use any physical restraints to restrict pelvic motion).
Our own data evaluating lumbar muscle activity during resistance exercise protocols has
also addressed this question. For example, we have published several studies where we
observed that during dynamic trunk extension without EPF, the EMG activity of the hip
extensors (gluteus maximum and biceps femoris) increased substantially with loading and
muscle fatigue suggesting that these muscles are highly responsible for accommodating the
required muscle force production [22-24]. One of these studies reported a decrease in lumbar
EMG activity occurred when subjects reached around 55% of their maximal fatigue capacity
(the total number of repetitions performed volitionally), and that a concomitant increase in hip
extensor activity resulted allowing for the continuation of the exercise task [23]. While this
fatigue-induced alteration in lumbar and hip extensor muscle activation patterns has not been
investigated under a condition of EPF it is likely that the synergistic contribution from the hip
extensors would be minimized and thus the task duration shortened. In a different study we
reported that when greater external loads were added to the lumbar extensors without EPF,
EMG amplitude increased to a greater extent in the larger gluteals and biceps femoris
musculature [22]. This finding is corroborated by findings from Mayer and colleagues
indicating that MRI signal intensity of the intrinsic lumbar extensor muscles does not increase
with an increase in exercise intensity from 50% to 70% of total trunk extension strength [25].
Collectively these findings suggest that when lumbar extension resistance exercise is
performed without external pelvic fixation the larger hip extensor muscles are more
responsible than the lumbar musculature for accommodating an increased load [22, 25].
SUMMARY
Compound lumbar-pelvic rhythm is responsible for trunk extension during normal daily
activities when the pelvis is not restrained [4,26,27,28]. Thus, one might argue that since EPF
does not occur in activities of daily living that performing more practical exercises that mimic
the activities of daily living, functional tasks or even pain management classes is a more
effective form of intervention [4, 36]. However, the scientific evidence suggests that this
approach primarily targets the larger, more powerful gluteal and hamstring (global) muscles,
and that the smaller, less leveraged, (specific) lumbar muscles remain in a relative state of
disuse and/or poor neural activation [5,29,31,32]. Conversely, when EPF is employed gains in
lumbar muscle strength and function are observed [4,5,6,10,11,12,13,30]. Neural activation
also appears to be enhanced as Graves and colleagues noted in their study [4]. They surmised
that the relatively large increases in strength, which resulted from only 12 exercise bouts was
most likely associated with a large neural component [4]. Similarly, Nelson and others found
that 94% of patients who received good or excellent results during a training study were able
to maintain their achieved training results even after a 1 year post-discharge [33]. Apparently,
the musculoskeletal dysfunction had sufficiently diminished allowing normal neuromotor
activity and functional tasks to be performed [34]. Certainly a large portion of the resistance
exercise training induced adaptations are a result of neural adaptations, and the role of
neuromotor control in both the pathogenesis and rehabilitation of low back pain has long been
of interest. A recent meta-analysis investigating the effectiveness of motor control exercises
in improving clinical severity of persistent low back pain concluded that motor control
exercise training alone is superior to minimal intervention, but that it is not more effective
than manual therapy or other forms of exercise [35]. Accordingly, they suggested a multi-
modal approach incorporating different exercise prescriptions (e.g., resistance exercise and
motor control training) along with manual therapies may be most effective in treating
persistent low back pain [35]. Accordingly, the functional relevance of successfully targeting
and training the lumbar extensor muscles via EPF (isolation) is suggested as an aspect of this
multi-modal approach.
REFERENCES
[1] McGill, S.M. Low back exercises: evidence for improving exercise regimens. Phys.
Ther. 1998; 78: 754-65.
[2] Bigos, S., O. Bowyer, and G. Braen, Acute low back problems in adults, P.H.S. US
Department of Health and Human Services, Agency for Health Care Policy and
Research, Editor. 1994, Clincal Practice Guideline No. 14, Rockville, MD.
[3] Choi, G., P.P. Raiturker, M.J. Kim, D.J. Chung, Y.S. Chae, and S.H. Lee. The effect of
early isolated lumbar extension exercise program for patients with herniated disc
undergoing lumbar discectomy. Neurosurgery. 2005; 57: 764-72; discussion 764-72.
[4] Graves, J.E., D.C. Webb, M.L. Pollock, et al. Pelvic stabilization during resistance
training: its effect on the development of lumbar extension strength. Arch. Phys. Med.
Rehabil. 1994; 75: 210-5.
[5] Pollock, M.L., S.H. Leggett, J.E. Graves, A. Jones, M. Fulton, and J. Cirulli. Effect of
resistance training on lumbar extension strength. Am. J. Sports Med. 1989; 17: 624-9.
[6] Carpenter, D.M. and B.W. Nelson. Low back strengthening for the prevention and
treatment of low back pain. Med. Sci. Sports Exerc. 1999; 31: 18-24.
[7] Fleck, S. and W. Kraemer, Designing resistance training programs 1987, Champaign,
IL: HUman Kinetics.
[8] O'Shea, P. Effect of selected weight training programs on the development of strength
and muscle hypertrophy. Res. Q. 1966; 37: 95-102.
[9] Mayer, J.M., B.E. Udermann, J.E. Graves, and L.L. Ploutz-Snyder. Effect of Roman
chair exercise training on the development of lumbar extension strength. J. Strength
Cond. Res. 2003; 17: 356-61.
[10] Carpenter, D.M., J.E. Graves, M.L. Pollock, et al. Effect of 12 and 20 weeks of
resistance training on lumbar extension torque production. Phys. Ther. 1991; 71: 580-8.
[11] Graves, J.E., M.L. Pollock, D. Foster, et al. Effect of training frequency and specificity
on isometric lumbar extension strength. Spine. 1990; 15: 504-9.
[12] Risch, S.V., N.K. Norvell, M.L. Pollock, et al. Lumbar strengthening in chronic low
back pain patients. Physiologic and psychological benefits. Spine. 1993; 18: 232-8.
[13] Holmes, B., S.H. Leggett, and V. Mooney. Comparison of female geriatric lumbar
extension strength: asymptomatic versus chronic low back pain patients and their
responses to active rehabilitation. . J. Spinal Disorders. 1996; 9: 17-22.
[14] Mooney, V., J. UGulick, and M. Perlman. Relationships between myoelectric activity,
strength, and MRI of the lumbar extensor muscles in back pain patients and normal
subjects. J. Spinal Disorders. 1997; 10: 348-356.
[15] Li, Y., L. Ploutz-Snyder, J. Graves, and J. Mayer. Neuromuscular adaptations to lumbar
extension strength gains. Med. Sci. Sports Exerc. 1998; 30: S207.
[16] Danneels, L.A., G.G. Vanderstraeten, D.C. Cambier, et al. Effects of three different
training modalities on the cross sectional area of the lumbar multifidus muscle in
patients with chronic low back pain. Br. J. Sports Med. 2001; 35: 186-91.
[17] Basmaijian, J.V. and C.L. DeLuca, Muscle Alive- their functions revealed by
electromyography. 1985, Baltimore: Williams and Wilkins.
[18] Adams, G.R., M.R. Duvoisin, and G.A. Dudley. Magnetic resonance imaging and
electromyography as indexes of muscle function. J. Appl. Physiol. 1992; 73: 1578-83.
[19] Yue, G., A.L. Alexander, D.H. Laidlaw, A.F. Gmitro, E.C. Unger, and R.M. Enoka.
Sensitivity of muscle proton spin-spin relaxation time as an index of muscle activation.
J. Appl. Physiol. 1994; 77: 84-92.
[20] San Juan, J.G., J.A. Yaggie, S.S. Levy, V. Mooney, B.E. Udermann, and J.M. Mayer.
Effects of pelvic stabilization on lumbar muscle activity during dynamic exercise. J.
Strength Cond. Res. 2005; 19: 903-7.
[21] Udermann, B.E., J.E. Graves, R.G. Donelson, L. Ploutz-Snyder, J.P. Boucher, and J.H.
Iriso. Pelvic restraint effect on lumbar gluteal and hamstring muscle electromyographic
activation. Arch. Phys. Med. Rehabil. 1999; 80: 428-31.
[22] Clark, B.C., T.M. Manini, J.M. Mayer, L.L. Ploutz-Snyder, and J.E. Graves.
Electromyographic activity of the lumbar and hip extensors during dynamic trunk
extension exercise. Arch. Phys. Med. Rehabil. 2002; 83: 1547-52.
[23] Clark, B.C., T.M. Manini, and L.L. Ploutz-Snyder. Derecruitment of the lumbar
musculature with fatiguing trunk extension exercise. Spine. 2003; 28: 282-7.
[24] Clark, B.C., T.M. Manini, and L.L. Ploutz-Snyder. Fatigue-induced changes in phasic
muscle activation patterns during dynamic trunk extension exercise. Am. J. Phys. Med.
Rehabil. 2007; 86: 373-9.
[25] Mayer, J.M., J.E. Graves, B.C. Clark, M. Formikell, and L.L. Ploutz-Snyder. The use of
magnetic resonance imaging to evaluate lumbar muscle activity during trunk extension
exercise at varying intensities. Spine. 2005; 30: 2556-63.
[26] Nelson JM, Walmsley RP, Stevenson JM: Relative lumbar and pelvic motion during
loaded spinal flexion/extension. Spine. 1995; 20:199-204.
[27] Neumann, Donald A, Kinesiology of the musculoskeletal system: foundations for
physical rehabilitation. 2002, St. Louis: Mosby.
[28] Farfan HF. Muscular mechanism of the lumbar spine and the position of power and
efficiency. Ortho. Clinic. N. Am. 1975; 6:135-44.
[29] Richardson, Carolyn., Hodges, Paul W., Hides, Julie, Therapeutic exercise for
lumbopelvic stabilization-a motor control approach for the treatment and prevention of
low back pain. 2nd Ed., 2004, Edinburgh: Churchill Livingstone.
[30] Deutsch FE. Isolated lumbar strengthening in the rehabilitation of chronic low back
pain. J. Manip. Physiol. Ther. 1996; 2:124-33.
[31] Hides JA, Richardson CA, Jull GA. Multifidus muscle recovery is not automatic after
resolution of acute, first-episode low back pain. Spine. 1996;21:2763-2769.
[32] O‘Sullivan PB, Phyty GD, Twomey LT, et al. Evaluation of specific stabilization
exercise in the treatment of chronic low back pain with radiological diagnosis of
spondylolysis or spondylolisthesis. Spine. 1997;22:2959-2967.
[33] Nelson BW, O‘Reilly E, Miller M, Hogan M, et al. The clinical effects of intensive,
specific exercise on chronic low-back-pain - a controlled-study of 895 consecutive
patients with 1-year follow-up. Orthopedics. 1995;18(10):971-981.
[34] Mooney V. Research justifies specific spinal training. Retrieved September 10, 2008,
from – http://www.biotech.com/db_area/archives/2000/0009sports.41-52.bio.html.
[35] Macedo LG, Maher CG, Latimer J, McAuley JH. Motor control exercise for persistent,
nonspecific low back pain: a systematic review. Physical Therapy. 2009;89(1):9-25.
[36] Critchely DJ, Ratcliffe J, Noonan S, et al. Effectiveness and cost effectiveness of three
types of physiotherapy used to reduce chronic low back pain disability: a pragmatic
randomized trial with economic evaluation [with consumer summary]. Spine.
2007;32:1474-1481.
Chapter 12
BONE CELL ADHESION: AN IMPORTANT ASPECT OF

CELL BIOMECHANICS IN THE DEVELOPMENT OF
SURFACE MODIFICATIONS FOR ORTHOPAEDIC
IMPLANTS
Andreas Fritsche1, Frank Luethen2, Barbara Nebe2,

Joachim Rychly2, Ulrich Lembke3, Carmen Zietz1,
Wolfram Mittelmeier1 and Rainer Bader1
1
Department of Orthopaedics, University of Rostock, Medical Faculty, Germany
2
Department of Cell Biology, University of Rostock, Medical Faculty, Germany
3
DOT GmbH, Rostock, Germany
ABSTRACT
Most revisions of total joint replacements are due to implant loosening, which is
mainly caused by wear particles (―wear disease‖) and inadequate primary implant
stability. The optimised integration of cementless total hip and knee endoprostheses into
the bone stock is the most adequate approach to achieve secondary implant stability and
to prevent implant loosening. Secondary stability is characterized by bone ingrowth of
the implant and decreases the amount of relative implant motion between the implant and
bone stock. It has also been suggested that prostheses which are fully occupied by bone
cells are less susceptible to infection. The economic impact of implant loosening is
immense, hence orthopaedic implant manufactures refine their products continuously.
Many technical developments have improved the survival rate of endoprosthetic
implants. Modern materials and surface modifications such as coatings help to reduce
wear rates, promote cell ongrowth or prevent infections. The cell adhesion of bone cells
onto implant surfaces has not been thoroughly investigated so far. However, different
methods to measure cell adhesion have been described. Some workgroups investigate
short-term adhesion or proliferation of bone cells on implant materials in-vitro, but little
is known about the long-term adhesion. Proliferation or short-term adhesion cannot

Tel.: +49 (0)381 494 9337, Fax: +49 (0)381 494 9308, Email: rainer.bader@med.uni-rostock.de
306 Andreas Fritsche, Frank Luethen, Barbara Nebe et al.
predict how strong the bonding between bone and implant will be. In most cases, cost
intensive animal studies have to be performed in order to gain expressive data. Hence, it
is important to assess the bone cell adhesion forces in an adequate experimental setup in-
vitro.
The exploration of bone cell adhesion on surfaces of orthopaedic implants
encourages the development of bio-compatible, bio-active and anti-infectious surfaces.
We have developed a test device, based on the spinning disc principle, which allows
quantitative measurements of osteoblastic cells on implant surfaces. First results show
differences in adhesion forces depending on the substrate. In future assessments different
bio-active and anti-infectious surface modifications will be analyzed regarding bone cell
adhesion prior to animal studies.
INTRODUCTION
Today, approximately 10 % of all total hip and knee endoprostheses need to be revised
after a 10 year period [1,2]. However, implant survival rates of more than 20 years have been
reported. The most common reason for implant failure is aseptic loosening mainly caused by
particle wear (―wear disease‖) [3,4]. According to Malchau et al. [1] almost 75 % of all hip
endoprostheses revisions are due to aseptic loosening. Other complications such as infections
(7 %), dislocations (6 %), fractures (5 %) [1] or implant allergies [5,6] occur less often but
can also lead to failure of the implant.
Insufficient primary and secondary implant stability can compromise the survival rate of
cementless implants [7]. Primary stability is the initial, post-operative fixation strength with
which the endoprostheses is held into place due to a press-fit force exerted by the bone.
Secondary stability is characterized by bone ingrowth of the implant and decreases the
amount of relative implant motion between the implant and the bone stock. The optimised
osseous integration of cementless hip and knee endoprostheses is the most adequate approach
to achieve secondary implant stability and to prevent implant loosening [8]. It has also been
suggested that endoprostheses which are fully occupied by bone cells are less vulnerable to
infection [9].
Besides the risk that patients have to take undergoing surgical treatment, the economic
impact of implant revision is immense. About 79,100 total hip and knee arthroplasties were
revised in the United States in the year 2005 [10], which amount to treatment expenses and an
economic damage of approximately 2.4 billion US-Dollars [11] per year. Infections
associated with joint replacement result in 0.36 billion US-Dollars of medical and surgical
treatment expenses annually [11].
In total joint replacement much effort has been put into the increase of implant survival
rates and hence, the reduction of implant revisions due to aseptic loosening. Since medical
implant materials are very demanding regarding biocompatibility and mechanical strength
[12,13] the variety of possible materials is highly limited. Therefore, the main emphasis for
endoprostheses improvement focuses on surface modifications. Coatings are already
successfully applied on implant-tissue interfaces in order to increase wear resistance [14],
promote osseous integration or to reduce the release of allergenic ions [6] (eg. Ni2+, Co2+,
Cr+). A commonly used surface coating for wear reduction as well as lowering allergenic ion
release is the titanium nitride (TiN) coating [15]. Newly developed coatings such as calcium
phosphate (CaP) [16] or plasma polymerised allylamine (PPAAm) [17] strive to improve
Bone Cell Adhesion 307
bony ongrowth, while others attempt to reduce infections [18] eg. by integrating copper in
titanium dioxide (TiO2-Cu2+) [19,20] or biodegradable PDLLA (poly d, l-acid) with
antibiotics [21].
In general, cell interaction with the implant interface plays an important role in surface
coating development. Cell spreading [22] and proliferation tests are carried out to acquire
biocompatibility data of new surface modifications [23]. But proliferation or even short-term
adhesion assays [24] cannot predict how strong the bonding between bone and implant will
be. Extra cellular matrix (ECM) proteins such as collagens, elastins, proteoglycans, and
adhesion proteins such as fibronectins and laminins [25] bind to a biomaterial surface within a
few minutes and enable cells to adhere to the implant surface. So far, expressive data of bone
cell adhesion on orthopaedic implants can only be obtained in cost intensive animal studies.
Hence, it is useful to assess the bone cell adhesion forces in vitro in an adequate experimental
setup to predict bone ongrowth and accompany the early stages of implant surface
modification development.
ADHESION ASSAYS
Different experimental setups to measure cell adhesion have been proposed. The simplest
method is the ―wash off‖ principle (Figure 1a) for short-term adherent cells (less than 60 min)
[26]. Non-adherent cells are washed off from the implant surface and can be counted
subsequently. This assay has no means of adhesive strength quantification. However, by
applying centrifugal forces in a centrifuge (Figure 1b) the adhesion forces can be determined.
This low-cost and easy to operate test setup [27,28,29,30] is only applicable to single cells or
cell populations (biofilm) that have bonded to the substrate for a short period of time, since
the utilised forces are not sufficient enough to detach long-term adherent cells. Nonetheless,
bonding strength of long-term, as well as short-term, adherent cells can be determined using
micromanipulative assays such as micropipettes (Figure 1c), laser tweezers or atomic-force-
microscope (AFM) cantilevers (Figure 1d) [30,31,32,33,34,35,36]. Normal and tangential
forces can be applied in order to detach adherent cells from the substrate. Although, results
are precise, only single cell measurements are possible and assessment of biofilms is
impossible using these techniques and apparatus which are cost-intensive and require
specialised operating personnel.
Measurements of short-term as well as long-term adhesive single cells or of cell
populations can be achieved using hydrodynamic flow assays. The relatively simple test setup
is based on a flow chamber in which a fluid flows over adherent cells on a smooth surface. At
a given flow pattern the generated shear stress w (force/area), respectively the resulting shear
force, that is exerted on the cells can be calculated. Especially at high flow rates it is required
to describe the flow patterns exactly since turbulences can arise and distort the results.
Furthermore, the shear force depends on cell size and shape [37,38]. Hydrodynamic flow
assays are classified by the type of flow generation or the build-up of the flow chamber. It can
be distinguished between linear flow between two parallel plates (Figure 1e), which has been
deployed successfully for cell adhesion [39,40], radial flow between two fixed, parallel
circular discs (Figure 1f) and flow in the laminar layer on the surface of a rotating disc
(spinning disc principle) [26].
Figure 1. Cell adhesion assays a) ―wash of‖ principle, b) centrifuge, c) micropipette, d) atomic force
microscope cantilever, e) linear flow between two parallel plates and f) radial flow between two fixed
parallel circular discs.
APPROACH TO MEASURE BONE CELL ADHESION FOR ORTHOPAEDIC

IMPLANT SURFACES
Since little is known about long-term cell adhesion forces of bone cells on implant
surfaces relevant for orthopaedic applications, we have developed an experimental adhesion
measuring device based on the spinning disc principle [26]. The spinning disc principle
describes a rotating disc in a resting fluid where the applied shear stress w = r·3/2 depends
on the rotational velocity  and the radial distance r from the centre of the disc (Figure 2).
Furthermore, the shear stress depends on the density  and the viscosity  of the fluid used
for the experiments. Laminar flow conditions on the surface of the disc even at high rotational
velocities are convenient in the attempt to measure cell adhesion. Another advantage is that
forces can be applied under constant uniform chemical and biological conditions since the
disc can rotate in cell culture medium [30].
Our measuring device consists of a stainless steel (V2A) shaft, a fluid chamber made of
polyetheretherketone (PEEK), a glass bottom, an electric motor and a disc shaped test sample
(Figure 3).
Figure 2. Spinning disc principle.
Figure 3. Adhesion measuring device adapted to be used in conjunction with a LSM shown
schematically.
The apparatus was designed to be placed on an inverse confocal laser scanning

microscope (LSM) so that cells stained with a fluorescent dye can be observed through the
glass bottom before and after rotation.
The test setup was built in such a way that the shaft can be moved towards the glass
bottom for microscopy and away for disc rotation. It is important to be far away from the
bottom of the fluid chamber during rotation in order to avoid interference of the arising
convection current [41] with the laminar boundary layer at the disc surface. The test
specimens were marked by a laser at different radial positions, so that distinct radial positions
could be retraced by microscopy.
First experiments were carried out using MG-63 cells, a human osteosarcoma cell line
(CRL-1427, ATCC, Rockville, Maryland, USA), with polished disc shaped test specimens
made of titanium alloy (Ti6Al4V) as well as Ti6Al4V samples coated with PPAAm. The
bone cells were stained with PKH26 (Sigma-Aldrich, St. Louis, MO, USA) and seeded onto
the specimens in 2 µl droplets of Dulbecco's Modified Eagle Medium (DMEM) with 10 %
fetal calf serum (FCS, Gibco Invitrogen, Karlsruhe, Germany) and 1 % Gentamicin
(Ratiopharm, Ulm, Germany) at defined radial positions. The cells were incubated at 37 °C
and 5 % CO2 and allowed to settle for 20 min before the samples were fully covered by
medium.
After a further incubation period of 18 h the test specimens were fixed onto the shaft of
the adhesion measuring device and the fluid chamber was filled with DMEM without FCS.
Subsequently, the apparatus was mounted onto the LSM (LSM410, Zeiss, Oberkochen,
Germany) (Figure 4).
The disc was rotated in the fluid for a period of 3 min at a given speed after moving the
sample away from the glass bottom. Imaging of the radial positions was performed before and
after rotation at the marked radial positions (Figure 5).
Figure 4. Adhesion measuring device mounted on LSM.

Figure 5. Miscoloured display of MG-63 after rotation showing adherent cells before (red) and after
rotation (yellow) at different radial positions.
The resulting shear stress res for the spinning disc at the radial position, where 50 % cell
detachment was detected, was calculated using equation 1.

η res  0.79981  r ηρω 3  (equation 1)
The obtained results show that the adhesion measurements are reproducible and
comparable to bonding strengths found in literature [30,42]. Moreover, a significant increase
in cell adhesion strength of MG-63 cells on a PPAAm surface coating compared to a polished
Ti6Al4V surface was detected. Animal studies show similar result outcomes. Further
exploration of bone cell adhesion on bio-active and anti-infectious orthopaedic implant
surfaces is currently in progress.
CONCLUSION
The economic impact of orthopaedic implant revisions due to implant failure is immense
which encourages further improvement of endoprostheses survival rates. Because potential
biocompatible materials are limited for orthopaedic applications, surface modifications have
become the focus of modern implant research. For higher survival rates sufficient
osseointegration (i.e. secondary implant stability) is essential. In-vitro cell adhesion tests
could indicate the potential of newly developed implant surface modifications regarding bone
ongrowth in early stages of the development.
To gain expressive test results different cell adhesion assays can be used, each with
advantages and disadvantages. We have decided on a rotating disc in a resting fluid (i.e.
spinning disc principle). This method is sensitive enough to show differences in cell adhesion
between different types of surface modifications. The results are reproducible and comparable
to those obtained by other workgroups. However, comparison and interpretation of the results
among different workgroups has to be performed carefully, since cell adhesion depends on
many influencing factors. Different cell types adhere with different strengths but also the
specimen‘s material, adhesion time period (i.e. incubation) or experimental parameters alter
the outcome of the results.
Further measurements on different bio-active, anti-infectious and anti-adhesive surfaces
and materials are being carried out. These investigations include the analysis of cell adhesion
time and different bone cell types (primary osteoblasts, mesenchymal stem cells).
Furthermore, the influence of metal ions (e.g. Cu2+) and the significance of specific integrins
on cell adhesion by blocking different integrin receptors will be studied extensively.
REFERENCES
[1] Malchau, H.; Herberts, P.; Eisler, T.; Garellick, G.; Söderman, P. J. Bone Joint Surg.
Am. 2002, 84, 2-20.
[2] Robertsson, O.; Knutsen, K.; Lewold, S.; Lidgren L. Acta Orthop. Scand. 2001, 72,
503-513.
[3] Zeiler, G.; Distler A. Biomaterialien 2000, 1, 19-24.
[4] Schuh, A.; Thomas, P.; Holzwarth, U.; Zeiler, G. Orthopäde 2004, 33, 727-731.
[5] Holzwarth, U.; Thomas, P.; Kachler, W.; Göske, J.; Schuh, A. Orthopäde 2005, 34,
1046-1051.
[6] Schuh, A.; Thomas, P.; Kachler, W.; Göske, J.; Wagner, L.; Holzwarth, U.; Forst, R.
Orthopäde 2005, 34, 327-328.
[7] Doehring, T. C.; Rubash, H. E.; Dore, D. E. Clin. Orthop. Relat. Res. 1999, 362, 230-
239.
[8] Widmer, K.-H.; Zurfluh, B.; Morscher, E. W. Orthopäde 1997, 26, 181-189.
[9] Gristina, A. G.; Naylor, P.; Myrvik, Q. Med. Prog. Technol. 1988-1989, 14, 205-224.
[10] Kurtz, S.; Ong, K.; Lau, E.; Mowat, F.; Halpern, M. J. Bone Joint Surg. Am. 2007,
89(4), 780-785.
[11] Darouiche, R. O. N. Engl. J. Med. 2004, 350(14), 1422-1429.
[12] Steinau, H.-U.; Schackert, H. K.; Bauer, H. In Chirurgisches Forum 2007 für
experimentelle und klinische Forschung; Springer: Berlin, (DE); 2007.
[13] Thull, R. Orthopäde 2003, 32, 51-59.
[14] Cook, G.; Costerton, J. W.; Darouiche, R. O. Int. J. Antimicrob. Agents 2000, 13(3),
169-173.
[15] Neumann, H.-G.; Baumann, A.; Wanke, G.; Hamelynck, K. J.; Morlock, M. Key Eng.
Mater 2002, 218-220, 601-604.
[16] Finke, B.; Lüthen, F.; Schröder, K. et al Biomaterials 2007, 28, 4521-4534.
[17] Becker, P.; Neumann, H.G.; Nebe, B. et al J. Mater Sci. Mater Med. 2004, 15, 437-440.
[18] Fritsche, A.; Haenle, M.; Mittelmeier, W. et al Mat-wiss u Werkstofftech 2008, 39(9),
659-664.
[19] Heidenau, F.; Stenzel, F.; Schmidt, H. et al Biomaterialien 2001, 2, 19-24.
[20] Heidenau, F.; Mittelmeier, W.; Detsch, R. et al J. Mater. Sci. Mater. Med. 2005, 16,
883-888.
[21] Gollwitzer, H.; Thomas, P.; Diehl, P. et al J. Orthop. Res. 2005, 23(4), 802-809.
[22] Lüthen, F.; Lange, R.; Becker, P.; Rychly, J.; Beck, U.; Nebe, J.G. Biomaterials 2005,
26(15), 2423-2440.
[23] Nebe, J.G.; Luethen, F.; Lange, R.; Beck, U. Macromol. Biosci. 2007, 7(5), 567-578.
[24] Nebe, B.; Finke, B.; Lüthen, F.; Bergemann, C.; Schröder, K.; Rychly, J.; Liefeith, K.;
Ohl, A. Biomol. Eng. 2007, 24(5), 447-454.
[25] Stevens, M.M; George, J.H. Science 2005, 310, 1135-1138.
[26] Garcia, A.J.; Huber, F.; Boettiger, D. J. Biol. Chem. 1998, 273(18), 10988-10993.
[27] Chu, L.; Tempelman, L.A.; Miller, C.; Hammer, D.A. AIChE Journal 1994, 40, 692-
703.
[28] Reyes, C.D.; Garcia, A.J. J. Biomed. Mater Res. 2003, 67A, 328-333.
[29] Giacomello, E.; Neumayer, J.; Colombatti, A.; Perris, R. Biotechniques 1999, 26, 756-
758.
[30] Garcia, A.J.; Gallant, N.D. Cell Biochem. Biophys. 2003, 39, 61-73.
[31] Evans, E.; Berk, D.; Leung, A.; Mohandas, N. Biophys. 1991, 59, 838-848.
[32] Litvinov, R.I.; Shuman, H.; Bennett, J.S.; Weisel, J.W. PNAS 2002, 99, 7426-7431.
[33] McKeever, P.E. J. Reticuloendothel Soc. 1974, 16, 313-317.
[34] Prechtel, K.; Bausch, A.R.; Marchi-Artzner, V.; Kantlehner, M.; Kessler, H.; Merkel,
R., Phys Rev Lett. 2002, 89, 28-101.
[35] Shao, J.Y; Hochmuth, R.M. Biophys. 2002, 77, 587-596.
[36] Tozeren, A.; Sung, K.P.; Chien, S. Biophys. 1989, 55, 479-487.
[37] Hammer, D.A.; Laufenburger, D.A. Biophys. 1987, 52, 475-487.
[38] Xiao, Y.; Truskey, G. A. Biophys. 1996, 71, 2869-2884.
[39] Truskey, G.A.; Pirone, J.S. J. Biomed. Mater. Res. 1990, 24, 1333-1353.
[40] van Kooten, T.G.; Schakenraad, J.M.; van der Mei, H.C.; Busscher, H.J. J. Biomed.
Mater. Res. 1992, 26, 725-738.
[41] Piva, M.; Meiburg, E. Phys. Fluids 2005, 17, 63603.
[42] Ma, G. ; Cai, N. ; Tan, S. ; Mhaisaklar, P.S. ; Chian, K.S.; Chan, V.; Liao, K. J.
Biomech. 2006, 39 (Suppl 1), 226.
Chapter 13
THE DIFFERENCES IN BIOMECHANICAL PATTERNS

OF FAST MOTOR LEARNING OF CHILDREN AND
ADULTS
A. Skurvidas1, A. Zuoza1, B. Gutnik2 and D. Nash2

1
Lithuanian State Academy of Sport and Physical Education,
Kaunas, Lithuania
2
Unitec, Institute of Technology, Auckland, New Zealand
The aim of this work was to establish and compare the patterns expressed on the fast
model of motor learning of children and adults executing a fast and accurate task.
The acquisition of a new motor skill follows two distinct stages with continued practice:
first, there is an early, fast learning stage in which performance improves rapidly within a
single training session; later, there is a slower learning stage within the time period of several
sessions of practice (Ungerleider et al., 2002). Motor learning is characterized by a specific
set of changes in performance parameters (Flament et al., 1999; Kempf et al., 2001). These
changes occur gradually in the course of a learning period. While the decreases or increases in
these parameters have been documented in a variety of tasks, it remains to be determined
whether the time of fast learning is different for children and adults. Therefore the main aim
of this study was to establish if there are differences in reaction time, average and maximal
velocity, trajectory, and accuracy as well as the variability of these parameters during motor
learning. The tasks involved 5 series with 20 repetitions in each.
For this experiment, we used a standard Analyzer DPA-1 measuring dynamic parameters
of movements of different segments of the body. The subjects performed a complicated task.
They had to react to a target on the computer screen as quickly as possible and follow the
most accurate trajectory towards the target, stopping within it. We measured the specific
parameters: reaction time (RT), average velocity (Va), maximal velocity (Vm), path of
movement (S), the period of achievement of maximal velocity (Tv) and time from
achievement of Vm to reaching the target.

Email: dnash@unitec.ac.nz
316 A. Skurvidas, A. Zuoza, B. Gutnik et al.
We found a significant difference (P<0.05-0.001) in RT, Va, Vm and S between children

and adults. One-way ANOVA showed a significant effect of time (five series: quick learning)
in Vm and S of children (P<0.01) and in Va of adults (P<0.05). There was a significant
difference in the changes of Va, Vm and S between children and adults. It is remarkable that
S and Vm decreased significantly after the first series and did not change during the 2nd-5th
series while Va increased gradually during all the five series in adults.
This is the first study to show, that the time of fast motor learning in a speed-accuracy
task was different between children and adults. We showed that the level of accuracy was
highly improved by the children at the expense of quickness of reaction, while adults
improved only the average velocity of their performance. It might be considered that the fast
motor learning process is expressed more in children than in adults. It is well established that
children have different information-processing capabilities compared with adults (Wade
1976).
Besides, most of the indices of variability of performance changed more significantly in
children than in adults. The reduction in performance variability was achieved along with
increased accuracy in children It might be speculated that children are more prone to change
motor variables during the fast learning process than adults. Though the underlying
mechanism for the fast learning difference between children and adults has to be elucidated, it
might be speculated that feed-forward and error correction mechanisms, as well as the
development of internal models of movement, may be more changeable in children than in
adults during the fast learning process.
Chapter 14
APPLYING PRESSURE SENSORS AND SIZE

DIFFERENCES IN RUNNING SHOES FIT
MEASUREMENT
Y. L. Cheng and Y. L. Hong

Department of Sports Science and Physical Education
The Chinese University of Hong Kong, Hong Kong
INTRODUCTION
Background
Fit is one of the most critical factors affecting footwear comfort (Miller et al., 2000).
Blistering, chafing, bunions and pain may be the result of poor fitting shoes. In long run, it
may cause the foot skeleton deformity.
feet, shoes, and the subjective perception for the shoes selection. Traditional method in
easily by tape measure and devices like Brannock (The Brannock Device Co., Inc., USA).
However these are considered to be insufficient for good footwear fitting (Witana et al.,
2004). Furthermore, researchers were also encountering problem in quantifying fit as it is
rather subjective which was also suggested to be affected by shoe wearing experience such as
tightness and looseness of the shoes (Kouchi et al., 2005). Therefore researchers are exploring
new method in measuring footwear fit, both objectively and subjectively.
3D technology has been applied to quantify footwear fit. Nacher et al. (2004) used 3D
foot digitizing technology to predict fit by getting the foot geometrical characterization data
and subjective fit perception from the users. Witana and the colleagues (2004) applied foot
scanners in their study to get the 3D foot shape which was then post processed to obtain the
2D outlines, and the dimensional differences between the feet and shoe last were analyzed
with the subjective fit perception. The dimensional differences plot allowed the shoe lasts
designer to determine the critical position for fit and unfit. One limitation for the 3D
318 Y. L. Cheng and Y. L. Hong
technology is that feet and last alignment has to be done manually. The actual position of the
feet in the shoes could not be ensured.
Apart from sizing, pressure distribution is likely to be a valid measurement for footwear
fit. Dewan et al. (2004) placed pressure sensors between the foot-shoe interfaces for dynamic
pressure measurement. The authors suggested that the pressure might have the implication for
footwear fitting and they had also identified the feasibility of using pressure sensors for
footwear fit. This was the only study that applied pressure sensors for fit measurement.
However this study did not relate the subjective perception of fit with the loading and how
each particular anatomical position of foot contributed to the overall fitting of the footwear.
With the interest to explore this area, Gheorghiu and the colleagues (2004) proposed to
investigate the relationship between pressure distribution and subjective fit rating.
The advantage of the pressure sensor application in footwear fit measurement is that it is
the real time assessment of the foot – shoe interaction. Unlike 3D technology, manual
alignment of feet and last was not required. The loading of the pressure sensors reflects the
actual interaction of the feet and the shoes. The investigation of the loading at different foot
anatomical position by using pressure sensors and by relating it to subjective fit perception,
critical position for fit may be identified. It is also hoped that norm can be built to show how
certain parts of the foot and shoe contribute to the overall fit feeling. However, despite the
advantages of using pressure sensor, its result can only reflect the shoes tightness but not
looseness. It is one of the limitations of pressure sensors application in fit measurement.
Subjective perception of fit of footwear had been measured in many studies (Gheorghiu
et al., 2004; Kouchi et al. 2005; Witana, et al. 2004). However none of the questionnaires
used above had been proved to be valid and reliable. Only one project had been done in
measuring footwear comfort by Mundermann et al. (2002) that tried to validate a footwear
comfort questionnaire. They tried to develop a reliable method to assess footwear comfort
during running by the Visual Analogue Scales (VASs) (Mundermann et al., 2002). A protocol
including a control condition during running was used and the resulting intraclass correlations
between comfort ratings for repeated conditions were found to be high. The authors thus
concluded that using VASs with a control condition in the protocol provided a reliable
measure to assess footwear comfort during running.
Despite the importance of fit in footwear has been announced in a number of researches,
only a few studies have been done on relating the shoe-foot size difference to the subjective
fit measurement. For the pressure distribution, there is even no concrete finding on the
association of the objective pressure measurement of the shoe-foot interface and the
subjective fit rating. Therefore, it is worthwhile to explore the feasibility of the application of
these objective measurements in measuring the subjective fit rating.
The purpose of this study was to find out the relationship between three variables:
subjective fit rating, foot-last size difference and pressure distribution of foot-shoe interface.
It is hoped that through the study of the two objective independent variables: foot-last size
difference and pressure distribution, variances of subjective fit rating can be more thoroughly
explained.
Furthermore, although previous studies have applied fit questionnaire in measuring
subjective fit perception, none of the questionnaire has been proven reliable. Therefore
another purpose of the present study is to test the reliability of a footwear fit questionnaire
that applied VAS scale with the control condition (Mundermunn et al. 2002). The reliable fit
questionnaire developed is necessary in quantifying subjective fit perception. Such
Applying Pressure Sensors and Size Differences… 319
information will be very useful in the area of research, production and sales in footwear
industry.
Although footwear fit is of high importance, present devices available in measuring fit
are obviously inadequate. Results of present study will help introduce new technology into
this area.
The knowledge of different objective measures in predicting subjective fit rating is
especially useful for those who are unable to express their subjective feel towards the fit of
the footwear e.g. young children, diabetes patients and the elderly.
Operational Definition
Foot length the distance between pternion (the rearmost point of the heel) and the
midpoint of the maximum breadth of the ball cross section. The ball cross section is the
vertical cross section passing through MT (metatarsale tibiale: the most medially protruding
point on the head of the 1st metatarsal bone) and MF (metatarsale fibulare: the most laterally
protruding point on the head of the 5th metatarsal bone).
Foot circumference the circumference of the ball cross section. Ball cross section is the
vertical cross section passing through MT. Foot breadth the maximum breadth of the ball
cross section. Heel breadth the breadth of the heel measured perpendicular to the foot axis at
the 16% of foot length from pternion.
Shoe last is the solid form around which a shoe is constructed resulting the inside shape
of the shoe (Figure 1). And each shoe last is designed for a particular heel height, toe shape
and type of footwear. Visual Analogue Scale (VAS) is a measurement instrument that tries to
measure a characteristic or attitude that is believed to range across a continuum of values and
cannot easily be directly measured (Gould et al., 2001). VAS that was used in this study is a
150mm horizontal line, anchored by word descriptors that enquire footwear fit perception at
each end. The subject will mark on the line the point that they feel which represents their fit
perception of their current state.
Figure 1. Shoe last.

Unlike the usual VAS used in other studies in which the VAS score was determined by
measuring from the left hand end of the line to the point that the subject marks, the
measurement starts in the middle in this study. The middle point of the line represents the
optimum perception of fit.
REVIEW OF LITERATURE
Problem of Poor Fitting
Fit is one of the most critical factors affecting footwear comfort (Miller et al., 2000).
Blistering, chafing, bunions and pain may be the result of poor fitting shoes. In long run, it
may be the cause of foot skeleton deformity.
Research regarding footwear fit problem in elderly has been carried out, it was found that
among the 65 elderly being investigated, 47 (72%) of them had ill fitting shoes (a discrepancy
in length of more than half a British shoe size or more than one British width fitting, 7 mm). It
was also found in the study that incorrect shoe length was significantly associated with the
presence of ulceration and self reported pain (Burns, Leese and McMurdo, 2002). In addition,
ill fitting shoes were also associated with the presence of corns on the toes, hallux valgus
deformity and foot pain (Menz and Morris, 2005). Gender was a factor found to be associated
with the prevalence of foot problem.
A study conducted by Frey and the colleagues (1993) evaluated the foot size and shoe
preference, shoe comfort, shoe size and the presence of foot pain and foot deformities of 356
healthy women between 20 and 60 years of age. In that study, 80% of women reported
significant foot pain while wearing shoes. 76% of the women had one or more forefoot
deformities, with hallux valgus being the most common (71% of this group had hallux valgus,
50% had hammertoes, 17% had bunionettes, 13% had prominent metatarsal heads, and 4%
had miscellaneous deformities). The overall incidence of hallux valgus was 54%.
The same study also showed that 88% percent of the women were wearing shoes that
were smaller in width than their feet (average, 1.2 cm smaller). In the women who had no foot
pain (20%), the average foot-to-shoe discrepancy was only 0.56 cm. In those women without
any deformity (23%), the average foot-to-shoe discrepancy was 0.6 cm.
It is clear from the above study that the majority of women wore shoes that were small
for their feet and they would suffer resultant pain and deformities. The patients in the study
wore a shoe that was 1.3 to 2.5 cm smaller than the width of the forefoot in average. The
patients without pain or deformity wore a shoe that was only smaller than the forefoot by 1.3
cm or less.
Many shoes for females are simply scaled down versions of the same shoe for males. In
general, women‘s feet were in a smaller width to length ratio than male, and the heel is
narrower in comparison with the forefoot (Frey, Thompson and Smith, 1995).
Not only gender plays a role in footwear fit, different foot size is also influential. A
similar study conducted by Frey and colleagues (1995) found that women with larger feet
would have more pain and deformities than the control group in the study. It was suggested
that because women with larger feet might have more difficulties in finding shoe with proper
fit than those with smaller feet. Women with wider forefeet tended to have wider heels. As
foot length increased, both forefoot width and heel width increased but with the heel width
increased to a lesser extend. However, as shoe length increased, all the key internal
dimensions in the shoes were enlarged in a fixed proportion. Therefore in order for the
women with larger feet to find the shoes with heel counter which could grip the heel firmly,
smaller shoe should be selected and it might be too snug for the forefoot resulting in
deformity and pain.
A study conducted by Nixon and the colleagues (2006) examined the shoes size and
width, sex, presence of diabetes, neuropathy and foot wound, type and condition of shoes
usually worn in order to find out the relationship of poorly fitting footwear with and without
diabetic foot ulceration. 400 US veterans were recruited. Of this population, only 25.5% were
wearing appropriately sized shoes. Patients with diabetic foot ulceration were 5.1 times more
likely to have poorly fitting shoes than those without a wound. The authors thus suggested
that greater accuracy in shoe-foot match could be useful in reducing the risk of foot
ulceration.
―Diabetic foot ulcerations are the most common severe complications of diabetes‖.
Ulcers form in patients with diabetes because of a lack of sensation, coupled with repetitive
pressure forces (walking). One of the central tenets in reducing the incidence of ulcers is
pressure reduction through the use of appropriate shoes and insoles. Poorly fitting shoes may
account for a large proportion of diabetic foot ulcers and may also play a role in creating or
exacerbating other complications in people without diabetes.
However, the result of a randomized controlled trial conducted by Reiber et al. (2002)
seems to contradict the above wisdom. A two year randomized trial of 400 diabetes patients
was conducted.
Participants were randomly assigned to receive customized shoes or their usual footwear.
It was found that there were as many foot ulcers in specifically designed footwear compared
with ordinary footwear. In respond to this finding, Chantelau (2002) commented the ways
Reiber et al. (2002) analyzed the results. With a closer look to the result of the study, 9 out of
47 ulcer episodes occurred in well-fitted intervention shoes, compared to 19 out of 37 ulcer
episodes occurring in normal shoes and these results were highly statistically significant.
However these results were not reported by Reiber et al. (2002). Chantelau (2002) thus
elaborated the finding to be supportive of the rationale that fit of footwear was important
rather than oppositing it.
Chantelau and Gede (2001) investigated the foot size of the elderly people with and
without diabetes mellitus in German. The result showed that the feet of most elderly people
were broader than the ‗normal‘ footwear that was currently available on the market. The
authors thus suggested that a fixed breadth-to-length ratio for the footwear construction might
not be enough to fit all the individual characteristics.
Apart from being one of the factors of foot deformities, fit also appears to govern
comfort. If a pair of shoes were tight, the pressure produced between the foot shoe interface
may cause pain to the wearer. On the other hand, if a shoe was too loose, the friction caused
by the slippery effect would also produce the sense of discomfort. (Luximon, Goonetilleke
and Tsui, 2003)
Definition of Fit
However the importance of footwear fit, it is not well defined. It is suggested that ―proper
shoe fit requires shape or last design with proper toe depth and shape, proper instep depth,
proper heel width, and proper curve (flare) of the shoe‖ (Tremaine and Awad, 1998).
However, this definition does not take into the account of individual characteristic of the
human feet and it also does not point out the importance of matching between human feet and
the shoes. As it was stated ‗closeness of match of shoe shape to that of the human foot‖ was
one of the most important factors for comfort of fit (Hawes et al., 1994).
One of the most common methods in measuring the fit of footwear was by finding out the
size difference between foot and shoes. Menz and Morris (2005) applied this idea by
assessing the footwear and foot dimensions and to find out their length, width and area. Ill
fitting footwear was defined as having size smaller than the foot. Similar concept was applied
by Nixon et al. (2006), inappropriately sized footwear was defined as a shoe size at least one
full size too large or too small using a specific foot measuring device.
Tremaine and Awad (1998), on the other hand, did not regard smaller size of the shoes as
unfit. Instead they proposed a range for fit should be equal or no more than 0.25 inch less that
fore-foot width. Witana et al. (2004) proposed a linear regression to predict forefoot fit by the
summation of the lateral and medial maximum dimensional difference of the forefoot.
According to the equation, maximal dimensional difference should be about -8.5mm.
Goonetilleke (2003) summarized a few important factors for footwear fit measurement:
dorsal arch height (a measure of foot height), plantar arch height, foot flare, and the angular
orientation of the metatarsal; another combination of the foot factors are: length, flare, width
and height. Although there are many guidelines available, the ultimate goal of footwear
selection is to find out the right match between the foot and the footwear, as Janisse (1992)
suggested: ―proper fit is achieved when shoe shape is matched to foot shape‖. The selection
of proper fit however is based on the subjective feeling of the person. Thus quantifying the
subjective feeling of fit becomes one of the importance research questions in footwear
industry.
feet, shoes, and the subjective perception for the selection of shoes. Traditional method in
easily by tape measure and other device like Brannock. However these are considered to be
insufficient for good footwear fitting (Witana et al., 2004). Furthermore, researchers were
also encountering problem in quantifying fit as it is rather subjective and it is suggested to be
affected by shoe wearing experience (Kouchi et al., 2005).
Recommendation of Shoe Fit
Determining correct shoe shape and the right size is the primary component in achieving
proper shoe fit (Janisse, 1992). Many recommendations are available for footwear selection to
ensure fit.
As suggested by American Orthopaedic Foot and Ankle Society1, a shoe should conform
the shape of the forefoot. Apart from shape matching, Fit can be achieved by having proper
shoe length that allow extra one half inch in the front. This extra space helps prevent the toes
from banging against the shoes. Proper shoe width is also important. Shoes must be wide
enough in the front to allow the toes to extend straight ahead and adequate room across the
ball of the foot (Janisse, 1992). Metetarsal should be situated in the widest part of the shoe.
Lace up shoes are recommended as laces allow adjustability. The lacing stops the
unnecessary movement of the foot (Janisse, 1992).
Sizing
In an attempt to evaluate the relevant dimensions of feet which are necessary for three
dimensional foot modeling, Goonetilleke et al. (1997) measured fourteen foot dimensions of
31 male subjects. The result of factor analysis, suggested that length, width, height at midfoot
and forefoot, and toe dimensions are needed to be considered for proper footwear fitting and
for the foot modeling.
3D technology has been applied to quantify footwear fit. Nacher et al. (2004) used 3D
foot digitizing technology to predict fit by getting the foot geometrical characterization data
and subjective fit perception from the users. Witana and the colleagues (2004) applied foot
scanners in their study to get the 3D foot shape which was then post processed to obtain the
2D outlines, dimensional differences between the feet and shoe last were analyzed with the fit
perception. The dimensional differences plot allowed the shoe lasts designer to determine the
critical position for fit and unfit. One limitation for the 3D technology, the feet and last
alignment has to be done manually that the actual position of the feet could not be ensured.
With an interest to know the 3D foot shape data without using the foot scanner, other
technology likes generating 3D foot shape from 2D information was proposed (Luximon et al.
2005). In order to develop the two models suggested by the authors, foot information of 40
participants was used and the models were validated by using another set of 40 Hong Kong
male subjects. In the first method, foot height and foot length were used to predict the 3D foot
shape by recursive regression equations. In the second method foot profile and foot shape was
used to predict the foot shape. Both methods were found to be valid in foot shape prediction
and both of them were relatively cheaper than using the foot scanner to determine the foot
shape for custom footwear design.
Pressure Distribution
Apart from sizing, pressure seems to be a valid measurement of fit. DeLooze et al. (2003)
states that pressure distribution appears to be the objective measure with the most clear
association with the subjective ratings. Although this study was done on car seat, this idea
should be applicable to footwear as well. Dewan et al. (2004) placed pressure sensors
between the foot-shod interfaces for dynamic pressure measurement. The authors suggested
1
10 Points for a Proper Shoe fit, American Orthopaedic Foot and Ankle Society, http://www.aofas.org/
i4a/pages/index.cfm?pageid=3300
that the loading might have the implication for footwear fitting and had also identified the
feasibility of using pressure sensors for footwear fit. This was the only study that applied
pressure sensors for footwear fit measurement. However this study did not relate the
subjective perception of fit with the loading and how each particular anatomical position of
the foot contributed to the overall fitting of the footwear. With the interest to explore this
area, Gheorghiu and the colleagues (2004) proposed to investigate the relationship between
pressure distribution and subjective fit rating.
Subjective Fit
Subjective perception of fit of footwear had been measured in many studies (Gheorghiu
et al., 2004; Kouchi et al., 2005; Witana, et al., 2004). However none of the questionnaire
used above had been proved to be valid and reliable. Only one project had been done on
measuring footwear comfort by Mundermann et al. (2002). They tried to develop a reliable
method to assess footwear comfort during running by the Visual Analog Scales (VASs). It
will be discussed in greater detail below.
Footwear Comfort
Apart from its influential role in foot health, footwear fit also contributes much in
comfort. Comfort is one of the most important factors for the design of footwear. However as
comfort is a subjective feeling depending on individual differences and it is difficult to
measure directly. Many methods were used to quantify comfort. Borg scale was widely used
to evaluate the comfort perception of the subjects. 15-point, 10-point and 5-point scale were
used respectively in different studies (Milani et al., 1997, Miller et al., 2000, Jordan and
Bartlett, 1995). Ranking was also one of the methods used by investigators (Chen et al, 1994).
Mundermann et al. (2001) used a visual analog scale (VAS) to assess footwear comfort
and later, they tried to develop a reliable method to assess footwear comfort during running
by the VASs (Mundermann et al., 2002). A protocol including a control condition during
running was used and the resulting intraclass correlations between comfort ratings for
repeated conditions were found to be high. It was suggested by the authors that a control
condition should be used before each test condition in order to have reliable comfort
measurement. For long term comfort, assessments with more sessions (four to six) should be
used. It was concluded that VASs provide a reliable measure to assess footwear comfort
during running when a control condition was included.
Investigators were interested in identifying biomechanical variables to measure the
comfort rating. Plantar pressure distribution was one of those under investigation. It was
reported that high plantar pressure in the midfoot area and low pressure in the medial forefoot
and hallux area were related to better comfort during walking (Chen et al., 1994). In the
study, fourteen subjects were first asked to rank the insoles in the order of comfort after
walking and running on a treadmill in self paced speed. After then plantar pressure was
measured by inserting a pressure distribution measuring device in the shoes. With more
comfortable insole, the path of centre of force at the plantar surface showed a movement to
the lateral side. It was suggested that measurement of pressure distribution might be useful to
detect change of shoe comfort. Similar result was obtained by Jordan and Bartlett (1995) and
a conclusion of the use of pressure distribution at foot-shoe interface to be a tool to identify
causes of discomfort in footwear.
Fit between the foot and shoe on comfort assessment was also studied. Good fit was
reported to be the most critical factor for shoe comfort. The most critical foot dimensions
included hallux height, dorsum height and length, forefoot and posterior calcaneus width and
toe lengths. If the fitting is good, other factors like skeletal alignment became important.
Moreover, shoes comfort was related to the activity performed and so did shoes torsional
stiffness and cushioning (Chen et al, 1994).
METHOLOLOGY
Design
The measurement of this study consisted of three parts: Foot scanning, pressure
measurement and perceive fit rating of running shoes. Foot scanning was to collect the foot
shape data of the subjects and also the shoes lasts‘ information. Pressure measurement was
applied to collect the pressure data between the foot-shoe interfaces. Subjective fit
measurement of the running shoes was obtained by fit questionnaire in order to collect the
subjective rating of the fit perception of the subjects.
Subject
Fifteen physically healthy male were recruited as subjects. Their foot sizes were screened
to be within the range of 25.5-27.5 cm. All subjects were injury free at the moment of study
and no injury history to be resulted in abnormal gait. Experiment detail was explained and
subjects were required to sign the informed consent before participating in this study).
Instrumentation
Fit Questionnaire. A fit questionnaire composed of 12 questions by using the VAS scale
was used. The questions include 1) Overall fit 2) free space in front of toe 3) fit at the side of
forefoot 4) upper forefoot fit 5) overall forefoot fit 6) height of dorsal arch 7) height of plantar
arch 8) fit at the side of midfoot 9) overall midfoot fit 10) free space at the back of rearfoot
11) fit at the side of the rearfoot 12) overall rearfoot fit.
The questionnaire was first developed in English and then translated into Chinese. Both
versions were amended and approved by respected professions respectively. Reliability of the
questionnaire was tested by asking the subjects to rate the test shoes again in the second day.
Foot scanner. The InFoot foot scanner (I-Ware Laboratory Co., Ltd) (Figure 2) was used
to collect the shoe lasts and subjects‘ foot shapes data. It is a 3D foot scanner which consists
of eight progressive CCD cameras and four semiconductor laser projectors.
Figure 2. InFoot foot scanner.
Pressure Sensors. 16 flexible pressure sensors (FSA, Canada) (Figure 3) was used to test
the shoe-foot interface pressure in the experiment. 12 out of the 16 sensors was attached on
the subjects‘ foot landmarks. Pressure data while standing and walking will be collected with
the frequency of 10 Hz.
Running shoes. 5 pairs of running shoes with the size of Eur 41 which were identical in
appearance but were made by different shoes lasts were in this study (Figure 4). A pair of
running shoes (Gel-Kayano XI, AXIS) was used as the control shoes in this study.
Figure 3. FSA pressure sensors.

Figure 4. Running shoes with identical appearance.
Shoe Lasts. The shoes lasts of the 5 running shoes (Figure 5) were provided by the shoe
producing company. They were scanned in order to get the lasts‘ dimensions: Length, ball
girth circumference, breadth, instep circumference and heel breadth.
Figure 5. Shoe Lasts: lateral view (left top); frontal view (right top); medial lateral view (left).
Procedure
Subjects came to the Human Biomechanics Laboratory for the experiment. They were
required to sign the consent form and their anthropometric data including body height and
body weight were measured and recorded. They first completed the foot scanning test and
then the subjective fit test and pressure distribution test were completed in the same section.
Foot Scanning. Feet of the subjects were cleaned by alcohol prep pad and twelve markers
were attached on the following foot landmark positions: 1) The most medial point of medial
malleolus 2) Sphyrion 3) The most lateral point of lateral malleolus 4) sphyrion fibulare 5)
Navicular 6) Tuberosity of 5th metatarsalis 7) Metatarsale tibiale 8) metatarsale fibulare 9) toe
1 joint 10) toe 5 joint 11) head of 2nd metatarsal bone 12) tentative junction point (Figure 6).
Foot scanning was done of half weight bearing state. Subjects were required to put one
foot in the foot scanner during foot scanning while another foot on a weight. Both feet were
scanned(Figure 7 and figure 8).
Landmark positions of the lasts were identified and they were scanned by the foot
scanner to get the lasts‘ dimensions.
Figure 6. Foot scanning- landmark positions.

Figure 7. Participants standing posture while foot scanning.
Figure 8. Half weight bearing while foot scanning.

Figure 9. Setting for shoes fit assessment. Participants were asked to walk on a 7 meter long route for 6
times before filling in the questionnaire.
Fit Questionnaire. Subjects were asked to evaluate the tightness or looseness of the
running shoes by filling the fit questionnaire which was designed by using a visual analogue
scale (VAS). There were totally 12 questions in the questionnaire. Fit at forefoot, mid foot
and rear foot position of the footwear were ranked. Detail instructions were given to each
subject to make sure every subject understand the questions clearly.
Set-up. Both subjective rating of standing and walking of the running shoes were
obtained. For the standing rating, subjects were required to stand for 30 second and then fill in
the questionnaire. They were then asked to walk on a 7 meter long route for 6 times before
filling in the questionnaire again (Figure 9). Such route was set up to make sure they have
completed a minimum walking distance before assessing the shoes‘ fit. Subjects were allowed
to fill in the questionnaire while walking.
Subject would first assess the fit of the control shoes and followed by the assessment of
each of the test shoes which were randomized in order. Between each trial of the test shoes,
subjects were requested to wear the control shoes and walked on the route for 3 times again in
order to make sure that they had the same fit sensation baseline for comparison.
Reliability Test In order to assess the reliability of the questionnaire, subjects were
required to come to the laboratory in the other day and to fill in the fit questionnaire again.
The sequence of the running shoes were randomized and same protocol as above was used.
Reliability for each shoed condition would be obtained to assess the repeatability of the
questionnaire.
Fit rating on the first day would be used for data analysis later. The rating got in the
second day was used to test the reliability of the questionnaire only.
Pressure measurement. Twelve of the sixteen flexible sensors were attached on the right
foot of the subjects on the following location: 1) Tip of toe 1 2) Metatarsal tibiale 3) Tip of
toe 2 4) Toe 1 Joint 5) Toe 5 joint 6) Metatarsal fibulare 7) Instep height 8) Navicular 9)
Tuberosity of 5th metatarsalis 10) Pternion 11) Medial calcaneous12) Lateral calcaneous
(Figure10, figure 11 and figure 12).
Figure 10. Pressure sensors location.
Figure 11. Anterior view of pressure sensors attachment (left).

Figure 12. Lateral view of pressure sensors attachment (top right).
Figure 13. Pressure data collection with the test shoes while standing (right).
Pressure data of three different conditions were collected: 1) socks only when standing 2)
test shoes when standing (Figure 13) 3) test shoes when walking. In condition 1, subjects
were required to wear the socks and stand with their weights balance at the center. Pressure
between the foot shoe interface were collected for 5 seconds. Similar to condition 1, subjects
stood steadily with the test shoes on in condition 2. In condition 3, pressure data were
collected while the subjects were walking on the route as in the questionnaire test (Figure 14).
Participants were asked to perform 3 trails in condition 1 and condition 2; 2 trials in condition
3.
When pressure data collection began, red light on the FSA data collection box flashed.
Video were taken to identify the red light and each heel strike during the walking trial. The
video was synchronized with pressure sensors.
Figure 14. Setting for pressure measurement.
In order to ensure the same shod condition during pressure data collection and subjective
fit test, pressure sensors were attached on the subjects‘ foot before the subjective fit test.
Therefore right after fit questionnaire completion, pressure data of the same running shoes
condition could be collected. Due to the thinness of the pressure sensors, the interference that
would be caused to the subjective fit perception was believed to be very small.
Data Reduction
Foot Scanning. In order to describe the fit of the shoes, size differences between the foot
and shoe last were found. Length, Ball girth circumference, foot breadth, instep
circumference, heel breadth the foot and last were obtained and five fit indicators were
calculated (figure 15) (Kouchi et al., 2005): 1) Foot length allowance (FLA): FL of last – FL
of foot; 2) Ball girth circumference allowance (BGCF): BGC of last – BGC of foot 3) Foot
breadth allowance (FBA): FB of last – FB of foot; 4) Instep circumference allowance (ICA):
IC of last –IC of foot]; 5) Heel breadth allowance (HBA): HB of last – HB of foot.
Fit Questionnaire. For each question, mid point of each line was defined as zero. Distance
between the mark and the mid point was measured. When the mark is on the left hand side of
the mid point, it was defined as negative while if it was on the right hand side of the point, it
was defined as positive.
Data of each question was then input into SPSS version 12 for analysis.
Video. Video taken during the pressure distribution test was captured. The time at which
the red light on the FSA collection box flashed was identified and synchronized. Time at each
heel strike was recorded.
Pressure Distribution. Pressure data of the sensors were exported to excel file. For the
sock standing and shoe standing trials, mean pressure values for each sensor were recorded.
As there were 3 trails for each shod conditions (5 pairs of shoe, excluding the sock condition
which acted as the baseline value), and there were 16 sensors. The total number of pressure
data for each subject was (3 X 5 X16) = 240.
1. Length
2. Ball girth circumference
3. Foot breadth
4. Instep circumference
5. Heel breadth
Figure 15. Foot dimension measurement.
Mean pressures got in sock standing condition was the baseline value. All the other data
would subtract the ground value before further analysis.
Data Analysis
SPSS (ver 12.0) was employed to perform all statistical analysis. Pearson Correlation
Coefficient and Intraclass correlation coefficient were determined in the reliability test of the
fit questionnaire. Stepwise regression was used to evaluate the relationship between the
subjective fit perception (dependent variable) and objective measurement (independent
variable). Statistical significance was accepted at the 0.05 level of confidence.
RESULT
Subject Detail
15 young male with aged 22.27+ 2.05 years, height 175.3+ 3.23 cm and weight 68.03+
7.49 kg were recruited in the study (table 1).
Table 1. Summary of descriptive statistic of the subjects
Mean + SD Maximum Minimum
Age 22.27+ 2.05 19 26
Height (cm) 175.3+ 3.23 180.2 169.6
Weight (kg) 68.03+ 7.49 54.04 70.05

Questionnaire
In order to find out the reliability of the questionnaire, Pearson correlation coefficients of
fit ratings measured by the fit questionnaires for different conditions were found. Standing
condition and walking condition were combined and analyzed together and were represented
as overall condition. The Overall test-retest reliability of each question in the questionnaire
ranges from 0.624 to 0.787. All the correlations are significant (p < 0.01). The test-retest
reliability of each question in standing and walking condition ranges from 0.538 to 0.813 and
all the correlations are significant (p<.01). Test-retest correlation coefficient of every question
for all subjects during standing and walking were shown in table 2 and figure 16.
ICCs of fit rating in different conditions were also found (Table 3). The ICC was 0.841
for overall fit rating in overall condition. The ICC for overall condition is 0.755 – 0.879;
standing condition is 0.754 – 0.864; walking is 0.695 – 0.894.
0.9
0.8
0.7
0.6
0.5
Overall
r
0.4 standing
walking
0.3
0.2
0.1
0
1 2 3 4 5 6 7 8 9 10 11 12
Question
Figure 16. Pearson Correlation coefficient of fit ratings in different conditions (overall, standing,
walking).
Table 2. Test-retest Pearson correlation coefficients of fit ratings for different conditions (Overall, standing and walking)
Question 1 2 3 4 5 6 7 8 9 10 11 12
Overall .726** .719** .626** .673** .741** .624** .719** .766** .787** .719** .635** .712**
rtt standing .738** .683** .624** .667** .764** .699** .695** .751** .764** .692** .625** .626**
walking .717** .755** .638** .683** .728** .538** .741** .788** .813** .745** .650** .792**
**.
Correlation is significant at the 0.01 level (2-tailed).
*.
Correlation is significant at the 0.05 level.
Table 3. ICC of fit ratings for different conditions (Overall, standing and walking)
Question 1 2 3 4 5 6 7 8 9 10 11 12
Overall .841 .834 .759 .801 .844 .755 .820 .846 .879 .833 .775 .832
ICC standing .848 .810 .754 .797 .859 .799 .809 .845 .864 .812 .764 .767
walking .853 .858 .771 .807 .828 .695 .830 .852 .894 .852 .787 .883
Fit Rating
Optimal fitting scored zero, negative values reflect tightness whereas positive values
reflect looseness. Apart from the fore foot fit rating for shoe 3, all the other fit rating were
ranked as negative which means that the testing shoes were generally ranked as tight by the
subject as compared with control shoes(Table 4 and Figure 17). When compared with the
other shoes, shoe 3 got all the overall fit, forefoot fit, mid foot fit and rear foot fit rating
closest to zero (overall fit = -2.50; forefoot fit = 0.23; mid foot fit = -2.6; rear foot fit = -3.2).
As a result, shoe 3 was rated the fittest by the participants among the 5 pairs of shoes.
ANOVA revealed significant difference in the forefoot fit rating. Post Poc Test result
showed that there was significant difference between shoe 3 and shoe 4 in forefoot fit rating
(p = 0.05). The average forefoot fit rating of shoe 4 was -13.53 and that of shoe 3 was 0.23
respectively.
Table 4. Mean score of overall, forefoot, mid foot and rear foot fit rating
Shoe 1 2 3 4 5
Overall fit -8.07 -8.34 -2.50 -13.34 -8.73
Fore foot fit -8.4 -9.63 0.23* -13.53* -9.97
Mid foot fit -8.43 -4.97 -2.6 -9.2 -6.93
Rear foot fit -8.37 -10.93 -3.2 -9.83 -5.40
shoe
1 2 3 4 5
2
-2
Overall fit
-4
Fore foot fit
-6
fit rating Mid foot fit
-8
Rear foot fit
-10
-12
-14
-16
Figure 17. Magnitude of fit ratings of the five testing shoes.

Size Difference
Table 5 showed the descriptive statistic of the subjects‘ feet: foot length (262.08 + 4.06
mm); ball girth circumference (255.79 + 8.99 mm); foot breadth (105.14 + 3.84 mm); instep
circumference (105.14 + 7.76 mm); Heel breadth (67.13 + 3.48mm).
All the subjects with foot length smaller than the length of the shoe last. For foot breadth
size, except in one foot-last combination (subject 15 and shoe last 3), all the subjects have
foot breadth larger than that of shoe last. The average size difference between the foot and the
shoe lasts were summarized in table 6.
Heel breadth was found to be moderately related to forefoot fit (r = 0.545; p<0.01) and
rear foot fit (r = 0.467; p<0.01). Low correlations were found between foot length and rear
foot fit (r = 0.254; p<0.01), foot breadth and forefoot fit (r = 0.203; p<0.05), foot breadth and
rear foot fit (r = 0.370; p<0.01), instep circumference and forefoot fit (r = 0.2387; p<0.01),
instep circumference and rear foot fit (r = 0.324; p<0.01), heel breadth and mid foot fit (r =
0.2; p<0.05) (Table 7).
Table 5. Descriptive statistic of subjects’ feet
Foot Ball Girth Instep

Foot Breadth Heel breadth
length Circumference Circumference
Mean (mm) 262.08 255.79 105.14 255.60 67.13
SD 4.06 8.99 3.84 7.76 3.48
Table 6. Foot – last size difference
Minimum Maximum Mean Std. Deviation

Foot length (mm) 3.20 19.70 12.27 4.10
Ball Girth Circumference (mm) -36.40 5.40 -12.01 9.19
Foot Breadth (mm) -16.20 .10 -10.22 4.00
Instep Circumference (mm) -13.20 19.30 1.39 8.34
Heel breadth (mm) -11.70 3.90 -2.44 3.65
Table 7. Pearson Correlation Coefficients between foot last dimensional difference and
fit rating (forefoot fit, mid foot fit and rear foot fit)
Ball Girth Foot Instep

r Foot length Heel breadth
Circumference Breadth Circumference
Overall Fit .120 .253** .301** .284** .478**
Forefoot Fit .085 .171* .203* .238** .545**
Mid foot Fit -0.89 .118 -.032 .120 .200*
Rear foot Fit .254** .173* .37** .324** .467**
**
*
Pressure Distribution
Pressure at MF was the highest among the 12 foot landmark positions, which was
followed by Toe 5 Joint. Average pressures of Shoe 3 were found to be the lowest in the 5 test
shoes at 8 landmark positions Toe1, Metarsal Tibiale, Toe 2, Toe 1 Joint, Toe 5 Joint,
Pternion, Medial Cancanous, Later Cancanous.(Figure 18).
One way ANOVA was applied to analyze the pressure at different foot landmarks
position of the 5 running shoes. No significant difference was found between the test shoes.
Moderate correlation was found between subjective overall fit and Toe 1 joint (-0.422;
p<0.01), Toe 5 joint (-0.419; p<0.01) (Table 8).
16
14
12
10
Shoe1
Mean Pressure 8
Shoe2
Shoe3
6
Shoe4
Shoe5
4
0
Toe1 MT Toe2 Toe1J Toe5J MF IH Nav Meta 5 Pternion M Cal L Cal
Figure 18. Mean pressure at different foot landmark positions (psi) of the 5 running shoes.
Table 8. Pearson Correlation Coefficients between subjective overall fit and pressure
Foot Toe1 MT Toe2 Toe1J Toe5J MF IH Nav Meta Pternion M Cal L Cal
Landmark 5
r - - -.11 - - .251* .075 - -.237 -.292** - -

.205 .187 .422** .419** .201 .343** .264*
Regression
Three regression models were performed by using dimensional differences only, pressure
only, dimensional differences and pressure as independent variables to predict the subjective
overall fit in standing which is the dependent variable. Stepwise multiple regression was used
to find out the model with the greatest R square which indicates that larger variance of
subjective fit rating can be explained.
0.3
0.2
0.1
0
Toe1 MT Toe2 Toe1J Toe5J MF IH Nav Meta 5 Pternion M Cal L Cal
r -0.1
-0.2
-0.3
-0.4
-0.5
Figure 19. Pearson Correlation Coefficients between subjective overall fit and pressure.
Dimensional differences only:
Table 9. Summary of step-wise regression results by using dimensional difference

variables as predictors of subjective fit rating
Adjusted R Std. Error of the

Model Variables Entered R R Square Square Estimate
1 Heel breadth .478(a) .228 .217 13.21577
2 Ball Girth
.554(b) .307 .286 12.62078
Circumference
a. Predictors: (Constant), Heel breadth.
b. Predictors: (Constant), Heel breadth, Ball Girth Circumference.
2 models were run in the stepwise regression for dimensional difference. Heel breadth
joint was entered in the first model and the R square was 0.228 (adjusted R square = 0.217).
Ball Girth Circumference was added in the second model and the resulting R square increased
to 0.307. Summary of the regression was shown in table 9.
Table 10. Coefficients, t values and significant values of step-wise regression results by
using dimensional difference variables as predictors of subjective fit rating
Unstandardized Standardized
Model Coefficients Coefficients t Sig.
B Std. Error Beta
1 (Constant) -3.040 1.906 -1.595 .115
Heel breadth 1.938 .435 .478 4.454 .000
2 (Constant) 2.547 2.737 .930 .356
Heel breadth 2.002 .416 .494 4.810 .000
Ball Girth
.451 .165 .280 2.732 .008
Circumference
Table 10 shows the coefficients, t values and significant values of step-wise regression
results by using dimensional difference variables as predictors of subjective fit rating.
Below is the regression equation that uses dimensional difference as predictors of
subjective fit rating:
Subjective overall fit = 2.547 + 2.002 (Heel Breadth) + 0.451 (Ball Girth Circumference)
R2 = 0.307 (adjusted R2 = 0.286)
Pressure only:
2 models were run in the stepwise regression for pressure sensors. Toe 1 joint was
entered in the first model with the resulting R square to be 0.168. MF was added in the
second model and the resulting R square increased to 0.255. Summary of the regression was
shown in table 11. Both predictors were found to be significant in the analysis (table 12).
The regression equation for pressure sensors as predictor of subjective overall fit is:
Subjective overall fit = 0.597 (MF) – 2.725 (Toe1 Joint) – 10.285
R2 = 0.255 (adjusted R2 = 0.237)
Table 11. Summary of step-wise regression results by using pressure sensors variables
as predictors of subjective fit rating
Std. Error of
Model Variables Entered R R Square Adjusted R Square the Estimate
1 Toe1J .422(a) .178 .168 13.14
2 MF .505(b) .255 .237 12.59
a Predictors: (Constant), Toe1J.
b Predictors: (Constant), Toe1J, MF.
using pressure sensors variables as predictors of subjective fit rating
Standardized
Model Unstandardized Coefficients Coefficients t Sig.
B Std. Error Beta
1 (Constant) -2.043 2.044 -1.000 .321
Toe1J -2.620 .625 -.422 -4.189 .000
2 (Constant) -10.285 3.469 -2.965 .004
Toe1J -2.725 .600 -.439 -4.541 .000
MF .597 .208 .278 2.878 .005
Dimensional differences and Pressure:

When using both the dimensional differences and the pressure for prediction, a final
equation with a higher R square (R2 = 0.626) was obtained (table 13). Three of the five
dimensional differences were entered in the equation: heel breadth, Ball girth circumference
and foot length. For the pressure data, pressure of the MF (metatarsal fibulare) and M cal
(medial calcaneous) were entered in the equation (table 14).
Table 13. Summary of step-wise regression results by using dimensional difference and
pressure sensors as predictors of subjective fit rating
Variables Variables R Adjusted R Std. Error of

Model Entered Removed R Square Square the Estimate
1 Toe1J . .517(a) .267 .255 13.11210
2 MF . .629(b) .396 .376 12.00389
3 Heel breadth . .670(c) .449 .421 11.56204
4 Ball Girth
. .714(d) .510 .477 10.98635
Circumference
5 M_cal . .740(e) .547 .508 10.65697
6 . Toe1J .731(f) .534 .503 10.71514
7 Foot length . .791(g) .626 .594 9.68005
a Predictors: (Constant), Toe1J
b Predictors: (Constant), Toe1J, MF
c Predictors: (Constant), Toe1J, MF, H_breadth
d Predictors: (Constant), Toe1J, MF, H_breadth, BG_Circ
e Predictors: (Constant), Toe1J, MF, H_breadth, BG_Circ, M_cal
f Predictors: (Constant), MF, H_breadth, BG_Circ, M_cal
g Predictors: (Constant), MF, H_breadth, BG_Circ, M_cal, F_length.
The regression equation by using dimensional difference and pressure sensors as

predictor of subjective overall fit is:
Subjective overall fit = 22.811 + 0.443 (MF) + 2.62 (Heel Breadth) + 0.680(Ball Girth
Circumference) – 2.962 (M Cal) – 1.286 (F Length)
R2 = 0.626 (adjusted R2 = 0.594)

using dimensional difference and pressure sensors as predictors of subjective fit rating
Model Unstandardized Standardized

Coefficients Coefficients t Sig.
B Std. Error Beta
1 (Constant) .500 2.351 .213 .832
Toe1J -3.668 .771 -.517 -4.756 .000
2 (Constant) -10.700 3.781 -2.830 .006
Toe1J -3.817 .707 -.538 -5.396 .000
MF .777 .216 .359 3.602 .001
3 (Constant) -7.688 3.853 -1.995 .051
Toe1J -3.003 .761 -.423 -3.945 .000
MF .641 .215 .296 2.975 .004
Heel breadth 1.066 .444 .263 2.398 .020
4 (Constant) .620 4.760 .130 .897
Toe1J -2.354 .761 -.332 -3.092 .003
MF .481 .213 .222 2.262 .027
Heel breadth 1.462 .446 .361 3.275 .002
Ball Girth
.480 .176 .272 2.730 .008
Circumference
5 (Constant) 3.241 4.773 .679 .500
Toe1J -1.176 .917 -.166 -1.283 .205
MF .419 .208 .194 2.009 .049
Heel breadth 1.759 .454 .434 3.872 .000
Ball Girth
.505 .171 .286 2.957 .004
Circumference
M_cal -1.673 .771 -.240 -2.169 .034
6 (Constant) 4.558 4.687 .973 .335
MF .346 .202 .160 1.715 .092
Heel breadth 2.088 .377 .515 5.537 .000
Ball Girth
.569 .164 .322 3.463 .001
Circumference
M_cal -2.259 .625 -.324 -3.616 .001
7 (Constant) 22.811 6.422 3.552 .001
MF .443 .184 .205 2.406 .019
Heel breadth 2.620 .369 .646 7.108 .000
Ball Girth
.680 .151 .385 4.497 .000
Circumference
M_cal -2.962 .594 -.425 -4.985 .000
Foot length -1.286 .340 -.359 -3.781 .000
DISCUSSION
The purpose of this study was to find out the relationship between the three variables:
subjective fit rating, foot-last size difference and pressure distribution of foot-shoe interface.
To my knowledge, this is the first study that tried to use both size and pressure
distribution as the predictors of subject overall fit of footwear.
Subjective Fit Rating
Reliability of the Fit Questionnaire

Pearson correlation coefficient and ICC were applied to examine the reliability of the fit.
It was a usual practice for researchers to use Pearson Correlation to test the reliability of a
questionnaire. However, recent research has questioned using correlational methods as a
measure of test retest reliability. Correlation is good at comparing deviations from the mean
on two measurements, but it is not sensitive to changes in the means of the scores (Vincent,
1999). Therefore ICC was also used in this study to ensure the reliability of the questionnaire.
ICC of the questionnaire fell within moderate to high in both standing and walking condition.
Similar to the finding of Mundermann and colleagues (2002), the reliability of VAS and a
protocol with a control condition to assess footwear fit was good. The satisfactory reliability
of the present study allowed the application of the fit questionnaire for fit assessment in the
present study.
There were large variations in test retest correlation in fit assessment between individuals
(table2). As suggested in Mundermann et al‘s study (2002), it may be due to the low foot
sensitivity of some individuals. Different individual has different sensory threshold to
pressure stimuli of the foot. The foot shod contact causes an input signal into the body. The
input signal into the system will be modified according to the change in size of the shoes. If
the modified input signal is above threshold in some individuals but below the threshold in
some individuals, differences in fit perception will be result.
Another reason for that may be due to change of foot sizes at different time of the day or
in different day. The foot size of an individual is smaller in the beginning of the day than the
end of the day. Therefore it is understandable that the fit perception of the shoe change along
with the foot size.
Besides, it is possible that some individuals are initially not able to use VASs in a reliable
way (Mundermann et al., 2002). The repeatability was found to improve if they were allowed
to fill in the questionnaires in more sessions and it was suggested to use four to six
assessments sessions. However, as there were only two sessions included in the present study,
it is thus understandable that some individuals might have low repeatability in filling in the
VAS questionnaire.
To sum up, the overall satisfactory reliability justifies the use of the questionnaire as the
accurate measurement of subjective fit rating in the present study.
Fit Rating
Although significant difference was only found in forefoot fit rating between shoe 3 and
shoe 4, shoe 3 was rated the best in average in the overall fit, forefoot fit, mid foot fit and
rearfoot fit. The result showed that shoe 3 fitted the subjects best in this study. The negative
fit rating of the other test shoes reflected the general sense of tightness of the test shoes felt by
the subjects. Shoe 3, which scored less negative then the other test shoes showed that it was
rated as less tight by the subjects. When we looked at the shoe last information of shoe 3, it
was found to have the largest foot length, ball girth circumference, foot breadth and heel
breadth.
When we compared shoe 3 with shoe 4, the instep circumference of shoe 3 is 4.5mm
larger. The difference is the largest among the test shoes in term of instep circumference and
it accounts for significant difference found in forefoot fit rating between the test shoes.
The difference between last 3 and the other shoe lasts were within 3 mm in all the
dimensions but yet, the small differences were proven to have an impact on the subject
subjective fit perception. The result shows the sensitivity of the foot to the dimensions
changes of the shoes which was within a few millimeters.
Size
Foot Shape
Five foot shape parameters were examined in the study: foot length, ball girth
circumference, foot breadth, instep circumference, heel breadth. The average foot length and
foot breadth of the subjects in the present study were found to be 262.08mm and 105.14mm
respectively. Hawes et al. (1994) found that Japanese and Korean with foot length smaller
than 265.7mm had the foot breadth average of 102.5mm. For the North American with the
same range of foot length, they had the foot breadth average of 99.1mm. Chantelau and Gede
(2002) found that individuals with foot length between 263mm-265mm had the foot breadth
between 97mm-101mm. When compared the foot shape of the subjects in present study with
those found previously, subjects in this study had wider foot breadth in general.
Dimensional Difference
The average foot shoe-length difference was 12.27mm (3.2-19.7 mm). All the subjects
have the foot size smaller than that of the shoes. This result was expected to be due to the
selection criteria of the foot size of the participants, those who had foot size larger than that of
the last would not have been invited to the study.
The average shoes breadth and foot breadth discrepancy in the present study was -
10.22mm. The negative value indicated that breadth of the shoe were smaller than the feet.
The dimensional differences here was greater than that found by Menz and Morris (2005)
which was -7.27mm for outdoor footwear; but smaller than that found by Frey et al (1993)
which was 12mm. According to the recommendation made by Tremaine and Awad (1998), a
shoe with proper fit should be equal or no more than 6.25mm (0.25 in) less that the fore-foot
width. Therefore the shoes used in the present study might be regarded as too narrow or unfit
to the participants. In addition, ball girth circumference and heel breadth also showed a
negative discrepancy when compared the foot data with the shoe last.
There is no universal agreement of the definition for footwear fit. Although there are a lot
of guidelines available, most of which have not been validated. Menz and Morris (2005)
defined incorrect size as at least half a British shoe size too small or too large. Nixon and the
colleagues (2006), on the other hand, defined inappropriately sized footwear as a shoe size at
least one full size too large or too small by using a foot size measuring device. However, a
proper fit of footwear should also put subjective feel of fit into consideration. Therefore it is
not possible to assess footwear fit only by looking at the dimensional difference between the
shoes and the foot. One would expect to know the function of the shoes, the activity they
perform before making the conclusion.
Correlation with Subjective Fit

No substantial relationships were found between size difference at a particular region and
subjective overall fit because size differences at different foot regions contribute together to
the overall fit rating. In addition, overall fit perception is a subjective measure that involves
individual preference of the shoe tightness and looseness (Kouchi et al., 2005). That explains
the low correlation found between ball girth circumference, foot breadth, instep
circumference and the subjective fit rating(r =0 .253, 0.301 and 0.284; p<0 .01).
As a result, it is interesting to find a moderately strong correlation between heel breadth
(r = .478, P < .01) and subjective fit rating, which is greater than that of the other foot regions
mentioned above. Au and Goonetilleke (2007) reported the distributions of 186 responses of
the importance of fit rating at different foot regions of comfortable and uncomfortable shoes
ladies‘ dress shoes. The majority of fit ratings were in the MPJ region (25.3%) and Toe region
(21.0%). Rear-foot region accounted for 16.7% which was relatively less than the above
regions. The result of Au‘s study showed that people would rate the fit in MPJ region (foot
breadth) as more important than rear – foot region (heel breadth).
The different results found showed the disagreement between subjective measurement
and objective measurement of fit rating. One might perceive MPJ region as more important in
influencing the overall fit but in fact, the rear foot region related more to the overall fit rating.
This interesting finding calls for further investigation in the topic. One of the reasons
suggested here is the rigidity of the rear foot part of the shoes. In the present study, the fore
and mid part of the shoes allowed certain adjustment by the shoe laces. The rear part of the
shoes, on the other hand is not adjustable. After the adjustment of the shoe lace, one may
perceive the fore part of shoe as more fit. At a result, the correlation between the subjective fit
rating and the dimensional difference of the fore part of the shoes was reduced.
No significant relationship was found between the dimensional difference of foot – last
length and the subjective fit perception. This finding contradicted the wisdom of the
contribution of shoe length in footwear fit. The sizes of the subject‘s foot were controlled
within 25.5cm – 27.5cm and the actual foot sizes ranged between 25.51cm-27.06cm. Such
screening prevented extreme unfit condition caused by improper foot length thus reduces the
sensitivity in detecting the correlation.
In an attempt to find out how much dimensional difference can explain the variance of
subjective overall fit, stepwise regression was applied. Heel breadth and ball girth
circumference were input in the equation and the resulting R square was found to be small (R2
= 0.307). Only 30% of subjective fit rating can be explained by foot-last dimensional
difference which is small for satisfactory prediction.
This result was expected due to the absence of strong correlation between dimensional
difference and subjective fit rating. Contradicting the result of a similar study conducted by
Witana and colleagues (2002), regression equation with high R square (R2 = 0.8247) was
found. One of the reasons accounting for the difference might due to the manual alignment of
the foot and shoe lasts. In Witana and colleagues‘ study (2002), the foot outlines were aligned
with each of the respective last outlines using the heel centerline and the measured end
distance between shoe and foot. This alignment, however was not done in the present study.
Furthermore, it was suggested that the perception of fit was significantly affected by
experience (Kouchi et al., 2005). The personal preference of tightness of the shoes reduced
the predictive power of dimensional difference which results in making generalization of the
result difficult.
Pressure
Greater pressures were generally added on the lateral side of the foot as it was shown in
the larger average pressure applied at MF and Toe 5 Joint (Figure 17). This may be due to the
protruding shape of the metatarsal fibula and the toe 5 joint. The contact of those areas with
the shoe produced pressure. Thus it is reasonable to see the relatively smaller average
pressure in shoe 3 which is, in average larger then the other test shoe. However, despite the
observable variations of the pressure between the foot-shod interface, it wasn‘t great enough
to demonstrate significant difference between the shoes.
Most of the significant correlation of subjective fit rating and pressure were found to be
negative. The result implies that tightness of the shoes accompanies poor fit sensation of the
subject. The correlation is stronger at the position of Toe 1 joint and toe 5 joint. However,
opposite result was found at MF, where fit was directly related to pressure. The result showed
that sensation of tightness varied from location to location. One may prefer a pair of shoes to
be tight at certain position but not another. This result is in coherence with the finding of Au
and Goonetilleke (2007).
When examining forces and pressures at the foot-shoe interface using pressure as the
measuring device, one might suspect that the presence of the pressure sensors themselves
could influence the parameters that were being considered. The high test and retest correlation
of the subjective test measurement, however, provided a proof that the presence of pressure
sensors may not influence the subjective fit measurement.
Similar to dimensional difference, stepwise regression was applied to examine how much
variance of subjective overall fit can be explained by pressure. MF and toe 1 Joint were input
in the equation. The resulting R square was found to be 0.237 showing that around 23% of
subjective fit rating can be explained by foot-shod interface pressure measurement. The
smaller R square of pressure suggests that dimensional difference is more capable in
explaining subjective fit than that of pressure.
Pressure measurement, unlike dimensional difference which can provide real time
pressure data while footwear fit is being assessed. However, the incapability of pressure
sensors in measuring shoe looseness reduces their prediction power of fit rating. Pressure
sensors, in one hand is very sensitive in measuring shoe tightness when the foot-shod
interface is in contact. But on the other hand, once the foot is not in contact with the foot, the
pressure sensors fail to detect the extend of looseness.
Dimensional Difference and Pressure
In order to improve the predicting power of the objective measures to the subjective fit
measurement, both dimensional difference and pressure were input in the stepwise regression.
The resulting regression equation was found to have a R square of 0.626 showing that 62.6%
of subjective fit rating can be explained by using dimensional difference and foot shod
interface pressure together. Four of the five dimensional differences were entered in the
equation: heel breadth, Ball girth circumference, medial calcaneous and foot length. For the
pressure data, pressure of the MF was entered in the equation.
The predicting power of the objective measures have increased twice from using solely
dimensional difference (R2 = 0.307) or solely pressure sensor (R2 = 0.255) to using both
measures together (R2 = 0.626). Despite only one pressure value was included in the
regression equation.
The result of the present study shows that by using 2 objective measures: dimensional
difference and pressure sensors at the same time can improve the explained variance of
subjective fit rating.
Subjective Fit and Objective Measures
Footwear fit is a controversial topic which is under the interest in the field of sport
biomechanics. Researchers have tried to quantify footwear fit by many methods. The biggest
challenge we are facing is the difficulty in quantifying footwear fit as ‗right fit‘ or
compatibility is generally unknown in many circumstances. As stated in Witana and
colleagues (2004) study, foot-shoe fit depends on many factors. Time of day, activity
performed will also influence the fit perception. As a result, footwear purchased at some time
on one day may not be as fit at another time of the day. Therefore, the role of subjective fit is
always downplayed especially by the professional in sports biomechanics. However, due to
the important role that footwear fit is playing in footwear comfort (Mundermann et al., 2003),
foot health and sports performance and which subjective perception should be one of the
determining factors.
However, not everyone has the precision sense to judge the footwear fit e.g. diabetes
patients, elderly and children. It is necessary to develop reliable objective measures in order
to predict or to estimate the subjective footwear fit.
Previous studies provided a lot of guidelines in footwear selection in order to ensure fit.
Some researchers have also tried different objective measures like foot scanning, foot-shoe
dimensional difference, pressure sensors to find out footwear fit. More attention has been put
in subjective fit rating nowadays. Many studies are now undergoing to explore the feasibility
to use different device to explain subjective footwear fit (Dewan et al., 2004; Gheorghiu et
al., 2004; Nacher et al., 2004; Witana et al., 2004). Many of which are of preliminary state of
exploration and seldom did these studies try using two kinds of objective measures at the
same time. Although the subject group in this study was not the target group we are focusing
on, it should be treated as the preliminary study for the application of measuring methods.
Therefore, result of the present study is very encouraging and it also gives us insight in future
footwear fit research direction.
CONCLUSION
This study examined the relationship of subjective rating and objective measures of
footwear fit by using dimensional difference and pressure sensors. Stepwise regression was
applied to find out the amount of explained variance that objective measure(s) could account
for the subjective fit rating.
Result showed that using dimension difference alone accounted for 30.7% of subjective
fit and 25.5% by using solely pressure sensors. When both objective measures were entered in
regression, R square increased to 62.6%. This indicated using two objective measures at the
same time increased the predictive power of subjective fit rating.
Dimensional difference and pressure sensors were able to compensate the limitations and
restrictions of one another which resulted in an increased predicting power of subjective fit
rating.
REFERENCES
Au, E. Y. L. and Goonetilleke, R. S. (in press). A qualitative study on the comfort and fit of
ladies‘ dress shoes. Applied Ergonomics.
Burns, S.L., Leese, G.P., and McMurdo, M.E.T. (2002). Older people and ill fitting shoes.
Postgraduate Medical Journal, 78,344-36.
Chantelau, E. (2002). Letter to the editor: shoe-fitting, doesn‘t it really matter? Gerontology,
48, 408.
Chantelau, E. and Gede, A. (2001). Foot dimension of elderly people with and without
diabete mellitus – a data basis for shoe design. Gerontology, 48,241-244
Chen H., Nigg B.M., and de Konging J. (1994). Relationship between plantar pressure
distribution under the foot and insole comfort. Clinical Biomechanics, 9(6)335-41.
DeLooze M.P., Kuijt-evers L.F. M., and Dieen J.V. (2003). Sitting comfort and discomfort
and relationships with objective measures, Ergonomics, 49(10)985-997
Dewan, C., Pearsall, D., and Turcotte, R. (2004). Dynamic pressure measurement about the
foot and ankle. Proceedings of XXIII International Symposium on Biomechanics in
Sports, Ottawa, CA, 505.
Frey, C. and Roberts, N. E. (2002). Problem shoes, problem: what to tell women about
footwear-recommendations on diagnosis, treatment, prevention. Women‟s health in
primary care, 5(11)682-91.
Frey, C., Thompson, F., and Smith, J., Sanders, M., Horstman, H. (1993). American
Orthopaedic foot and ankle society women‘s shoe survey. Foot and Ankle, 14(2) 78-81.
Gheorghiu, C.R., Pearsall, D.J., and Turcotte, R.A. (2004). Quantifying fit in ice hockey skate
books. Proceedings of XXIII International Symposium on Biomechanics in Sports,
Ottawa, CA, 509.
Goonetilleke, R.S. (2003). Designing footwear: back to basics in an effort to design for
people. Proceedings of SEAMEC, Kuching, 25-31.
Goonetilleke, R.S., Ho, E.C.F., and So, R.H.Y. (1997-98). Foot sizing beyond the 2-D
Brannock Method. Annual Journal of the Institute of Industrial Engineers (Hong Kong).
28-31.
Gould, D., Kelly, D., Goldstone, L., and Gammon, J. (2001). Examining the validity of
pressure ulcer risk assessment scales: developing and using illustrated patient simulations
to collect the data. Journal of Clinical Nursing, 10, 697-706.
Hawes, M.R., Sovak, D., Miyashita, M., Kang S.J., Yoshihuku, Y., and Tanaka, S. (1994).
Ethnic differences in forefoot shape and the determination of shoe comfort. Ergonomics,
37(1) 187-96.
Janisse, D.J. (1992). The art and science of fitting shoes. Foot and Ankle, 13(5) 257-262
Jordon, C. and Bartlett R. (1995). Pressure distribution and perceived in casual footwear. Gait
and Posture, 3(4)215-220
Kouchi, M., Mochimaru M., Nogawa, H., and Ujihashi, S. (2005). Morphological fit of
running shoes: perception and physical measurements, Proceedings of the 7th symposium
on footwear biomechanics, Cleveand, OH, USA, 38-39
Luximon, A., Goonetilleke, R. S. and Tsui, K. L., (2003). Footwear Fit Categorization, the
Customer Centric Enterprise:Advances in Mass Customization and personalization,
edited by Mitchell M. T. and Frank T. P., Springer Verlag, pp 491-499.
Menz, H. B., and Morris, M.E. (2005). Footwear characteristics and foot problems in older
people. Gerontology, 51, 346-351
Milani T.L., Hennig E.M., and Lafortune M.A. (1997). Perceptual and biomechanical
variables for running in identical shoe constructions with varying midsole hardness.
Clinical Biomechanics, 12(5)294-300
Miller, J.E., Nigg B.M., Liu W., Stefanyshyn D.J., and Nurse M.A. (2000). Influence of foot,
leg and shoe characteristics on subject comfort. Foot and Ankle international, 21(9)759-
67.
Mundermann A., Stefanyshyn D.J., and Nigg B.M. (2001). Relationship between footwear
comfort of shoe inserts and anthropometric and sensory factors. Medicine and Science in
Sports and Exercise, 33(11)1939-45.
Mundermann A., Nigg B.M., Humble R.N., and Stefanyshyn D.J (2003). Foot orthotics affect
lower extremity kinematics and kinetics during running. Clinical Biomechanics, 18(3)
254-262
Mundermann A., Nigg B.M., Stefanyshyn D.J., and Humble R. N. (2002). Development of a
reliable method to assess footwear comfort during running. Gait and Posture, 16(1) 38-
45
Nácher, B., Alcántara, E. Alemany, S., García-Hernández, J., and Juan, A. (2004). A 3D foot
digitizing and its application to footwear fitting. Proceedings of 3D Modellling, Paris,
France.
Nixon, B.P., Armstrong D.G., Wendell, C., Vazquez J.R., Rabinovich, Z., Kimbriel, H.R. et
al.(2006). Do US Veterans wear appropriately sized shoes?: The veterans affairs shoe
size selection study. Journal of the American Podiatric Medical Association. 96(4) 290-2.
Jordan C., and Bartlett R. (1995). Pressure distribution and perceived comfort in casual
footwear. Gait and Posture, 3(4)215-20.
Reiber, G.E., Smith, D.G., Wallance, C., Sullivan K., Hayes, S., Vath, C. Maciejewski, M.L.,
Yu, O., Heagerty, P.J. LeMaster, J. (2002). Effect of the therapeutic footwear on foot
reulceration in patients with diabetes, Journal of American Journal Association, 287,
2552-2558.
Tremain, M.D. and Awad, E.M. (1998). The foot and ankle source book. Los Angeles:
Lowell House.
Vincent, W.J. (1999). Statistics in kinesiology. Champaign, IL : Human Kinetics.
Witana, C.P., Feng, J., Goonetilleke, R.S. (2004). Dimensional differences for evaluating the
quality of footwear fit. Ergonomics, 47(12)1301-17.
Editor: Jerrod H. Levy, pp. 353-385 c 2010 Nova Science Publishers, Inc.

Chapter 15
I MPROVEMENT OF THE I NPUT D ATA IN

B IOMECHANICS: K INEMATIC AND B ODY S EGMENT
I NERTIAL PARAMETERS
Tony Monnet1, Mickaël Begon2 , Claude Vallée1 , Patrick Lacouture1∗
1
Université de Poitiers, Laboratoire de Mécanique des Solides
UMR CNRS 6610, Bvd Marie et Pierre Curie BP 30179
86962 Futuroscope Cedex, France
2
Department of Kinesiology,
University of Montreal,
Research Centre of Sainte-Justine Hospital,
Montreal, Canada
Abstract
Usually, biomechanical models used for human motion analysis are oversimpli-
fied, especially for clinical analyses (e.g. Helen Hayes model). The calculated net
joint forces and torques are sensitive to the input data: segment kinematics and body
segment inertial parameters. It is therefore necessary to improve these input data us-
ing new methods and models adapted to the population and movement of interest. The
general problem is divided into three parts: (i) minimization of soft tissue artefacts,
(ii) joint centre location and (iii) identification of the personalized body segment pa-
rameters.
1 Introduction
In sports and clinical biomechanics, most of the analyses are based on the joint kinematics
and kinetics using force sensors e.g., force platform) and motion analysis systems. Since
the optoelectronical motion analysis systems measure only the spatial coordinates of targets
put on the skin, several steps — calculations and assumptions — are involved in the dynam-
ical analysis of the movement. In the eighties, three-dimensional (3D) kinematical models
∗
tony.monnet@univ-poitiers.fr
354 Tony Monnet, Mickaël Begon, Claude Vallée et al.
appeared for gait analysis using a few markers [28, 39] ; these models and their marker sets
have been progressively adopted as standard models by motion analysis software (e.g., Vi-
con Polygon and OrthoTrack from Motion Analysis Corporation) to automate the process
from the marker capture to the joint torque calculation and some interactive multimedia
reports.
Figure 1: Helen Hayes model in [28].
Helen Hayes model (Figure 1) for example was proposed with a minimal set of markers
and assumptions about the lower-limb joint centres. Fifteen markers are needed for captur-
ing the lower-limbs (seven body segments) and then calculating the kinematics and kinetics.
This model is still widely used for two main reasons: the markers are easy to track in a 3D
space with optoelectronical motion analysis systems [64] and a large database has been ob-
tained from this model. In gait analysis, one of the goals is to discriminate between normal
and pathological gaits and to assess changes in dynamics over time. Due to the error magni-
tude widely quantified by the literature [11, 17, 29, 38], these simplest models are suitable
for movie or video game motion capture. To some extent, they can identify intra-subject
changes between various conditions and large inter-subject or population differences, since
some assumptions (joint centre locations, inertia parameters) can affect differently two sub-
jects or two populations. In applied biomechanics or research fields using human movement
measurement tools (e.g., motor control), motion analysis systems combined with analysis
software are sometimes used as an ”intelligent black box” where the marker placement,
height and body mass are the input and the joint angles and torques are the output. How-
ever the relationship between marker coordinates and joint torques is not so obvious; it
Improvement of the Input Data in Biomechanics 355
appears indirectly in the motion equation. Using the Newton-Euler formalism proposed by
[5], the equations of motion for one segment of a serial chain read:

(1)
m1 (r̈P1 − g) + 10 R̈ m1 c1 = F, (1)
h i h i
(1) (1)
(r̈P1 − g) ⊗ m1 c1 10 RT − 10 R m1 c1 ⊗ (r̈P1 − g) +
1 (1) T (1) T
0 R̈ KP1 10 R − 10 R KP1 10R̈ = MP1 , (2)
the details of the notations are presented in the second part of this chapter (3.1).
The matrix of rotation (10R) is sensitive to the marker positions. It respects, as much
as possible, the anatomical axes of the segment to describe the joint kinematics and ki-
netics in line with the functional anatomy e.g. flexion-extension, abduction-adduction and
mediolateral rotations). Additionally, Equations 1 and 2 show that the joint efforts appear
to depend linearly on to the mass (m1 ), the first order moments (m1 c1 ) and the matrix of
inertia (KP1 ).
In parallel to the use and abuse of these models for any kind of movements and popu-
lations, modeling research and tools have progressed. Nowadays, with the improvement in
resolution of the motion analysis systems, increasing the number of markers is no longer a
problem. Hundreds of markers can be tracked and identified in real time with a high degree
of accuracy; medical imaging devices ( e.g. full-body CT scan, DXA) are used to esti-
mate the segment inertial parameters and the human body models become more and more
realistic, e.g. introduction of wobbling masses [52] or splitting the foot in subsegments
[58].
A first source of uncertainties is the skin movement artefact that affects the skeletal
kinematics. Recently, methods and protocols have appeared for estimating the soft tissue
artefact or the effect of wobbling masses by placing multiple markers on the body segments
(Figure 2a). In their paper, Stagni et al. [62] evidenced that soft tissue artefact is subject-
and task-specific, and that critical errors are propagated on knee kinematics. In a recent
paper, Monnet et al. [46] proposed a method for determining the most deformed markers
and the lesser one on the thigh (Figure 2b). A set of 26 markers was placed on the thigh and
the most deformed markers were iteratively eliminated until only three markers remained.
First results of this study highlight that the worst markers are located on the medial part of
the thigh (1-2, see Figure 2b) or close to joints: the greater trochanter (17, see Figure 2b)
and above the patella (10, see Figure 2b). These recent papers have a direct application for
the joint centre estimation using functional methods [32]. These methods consist in locating
the joint centre from the kinematics of the adjacent segments. Functional methods are not
commonly implemented and predictive methods are still used in many research studies and
in clinical settings. For human hip joints, predictive approaches have been shown to be
accurate to approximately 20 mm when compared with X-ray measurements [10, 50]. The
main purpose of the functional methods is to improve this accuracy whatever the marker
placement and the subject morphology.
The second source of uncertainties concerns the body segment inertial parameters. They
are commonly estimated using regression equations of proportional anthropometrical mod-
els determined from specific populations e.g. Bulgarian [51], Caucasian [68]). Recent
papers in this field aim at proposing new and more accurate approaches for estimating
356 Tony Monnet, Mickaël Begon, Claude Vallée et al.
[b] Determination of the most deformed markers [46].
[a] Quantification of soft tissue

artefact [62].
Figure 2: Redundant set of markers for [a] the quantification of soft tissue artefact or [b]
the determination of the most deformed markers.
subject-specific body segment inertial parameters. Most of the times, they consist in the use
of medical imaging [4, 42, 67] and cannot be used easily and widely for sport or clinical
analysis.
This chapter aims at describing recent techniques for optimizing the segmental kinemat-
ics and identifying subject specific inertia parameters. Both parts are introduced by a short
state of art to further explain the limitation of a standard protocol and present a range of
solutions proposed in the literature. This chapter is concluded by a few recommendations.
2 Kinematics
To analyse the 3D kinematics of the skeleton, local systems of coordinates should be as-
sociated with each segment. Local systems of coordinates are defined using anatomical
landmark and joint centre locations [19, 20] to interpret the joint kinematics according
to the functional anatomy. Most of the time, anatomical landmark positions are located
by placing reflective markers on the skin above bony prominences. These markers are
called anatomical markers while markers put elsewhere on the segment are called technical
markers. Another approach has been proposed for locating anatomical landmarks because
anatomical markers close to the joints are highly sensitive to soft tissue artefacts [18, 20]
(e.g. anatomical marker corresponding to the greater trochanter). This alternative method
is called the Calibrated Anatomical System Technique (CAST, [17, 19]): anatomical land-
marks are pointed using a wand composed of at least two markers with a known position
relative to the tip. The location of the anatomical landmark, i.e. the wand tip, may be de-
termined in the laboratory-fixed systems of coordinates and then in a corresponding local
system of coordinates using the technical markers. These technical markers can be put all
over the segment, specially far from joints and large soft tissues areas. Consequently their
kinematics are less affected by skin movement artefact [20].
Concerning the joint centres, they are located through predictive or functional methods.
The glenohumeral and the hip joint centre locations are far from the skin surface and thus
difficult to locate from bony landmarks. However they are of many interests because these
locations are required to define humeral and femoral local systems of coordinates. The pre-
dictive approach is based on regression equations: of the scapula geometry for estimating
the glenohumeral joint location, of the pelvis geometry for the hip joint location (Figure
3). Equations of regression can be found in Bell et al. [10] for the hip joint centre and in
Meskers et al. [45] for the glenohumeral joint centre. This approach is mainly affected by
PC BA
RPSIS LPSIS
AA
TS
LASIS
RASIS
SP
AI
Figure 3: Anatomical landmark positions for locating the glenohumeral joint centre (AA:
Angulus Acromialis, AC: AcromioClavicular joint, AI: Angulus Inferior, PC: Processus
Coracoideus, TS: Trigonum Spinae) and the the hip joint centre (ASIS: Anterior Superior
Iliac Spine, PSIS: Posterior Superior Iliac Spine, SP: pubis symphysis) from regression
equations.
errors on anatomical landmark calibration and by the regression uncertainties that are spe-
cific to a given population [27]. Furthermore reflective markers need to be stuck on specific
bony landmarks with accuracy. Hence for pathological subjects with skeletal deformations,
predictive approaches are not suitable. Among the functional methods (e.g. SCoRE [32],
SARA[33], Helical Axis[69]), none of these has definitively shown a superiority. Several
parameters affect the location like the type of the movement ( i.e. flexion/extension, abduc-
tion/adduction, circumduction) and the nature of movement (velocity and range of motion)
[7]. The work of Begon et al. [7] is based on experimental and simulated data. The simu-
lation is constructed from systematic and random noise model to reproduce skin movement
artefact that showed a cyclic behaviour with the largest deviations observed near the max-
imal flexion or abduction. The nature of the movements offers a compromise between a
variety of postures and the effect of skin movement. The variety of postures obtained by
using different types of movement is always positive in terms of accuracy (Figure 4).
Since in each study participants perform different movements, the comparison and rank-
ing of functional methods remain difficult. Of course, the precision and accuracy of these
approaches depend mainly on the quality of the kinematical data. A Recent paper haa
Figure 4: Graphical representation of the multiple comparison procedure of the nested

ANOVA. The standard error of the estimated means was calculated using Tukey’s post hoc
test. The nature of the movement, the nature-type interactions and the nature-type-cycles
interactions were tested successively. Two means were considered to be significantly differ-
ent when their intervals did not overlap, and were considered as not differing significantly
when their intervals overlapped (p < 0.01) [7].
shown that the Symmetrical centre of Rotation Estimation ( SCoRE) algorithm is better
than the Helical Axis method which is recommended by the ISB [48] because SCoRE al-
gorithm is not affected by slow velocities. Therefore its precision can be improved using
minimization method for determining the kinematics [8]. Much research has been done and
is still going on to improve the accuracy of functional algorithms that determine optimal
joint centre location. However, there is a lack of research in whether the obtained locations
can be used accurately and effectively when analysing subsequent human movement, spe-
cially where the noise from soft tissue deformation is substantially different from the trials
used to determine the optimal location. A few studies have been focused on this part of the
biomechanical analysis [57, 8].
Soft tissue artefact can also be reduced using optimization methods. Human kinetics
calculation is often based on multibody dynamics assuming pin joints without translation.
However with at least three markers, each body segment can be considered independently
of the proximal one and will have three degrees of freedom in rotation and three degrees of
freedom in translation. This results in apparent joint dislocation. To reduce this problem
and the skin movement artefacts, Lu and O’Connor [44] proposed a global optimization
procedure with a chain model, i.e. the degrees of freedom at each joint and the segment
lengths are previously defined from the joint centre locations. The positions of the reflec-
tive markers are calculated in the local frames of the corresponding segments and a global
optimization solves a multi-target inverse kinematics problem for the body configuration.
This method has been successfully applied to computer simulated movements of the lower-
limbs [44] and the upper-limbs [59]. Thereafter Begon et al.[9] have shown the interest
of this method for movement with unsolvable simulation and for occlusions. Global op-
timization provides a great opportunity to design optimal marker sets to minimize skin
movement artefact, because less than three markers are needed on each body segment and
the noisy markers can be removed. This global optimization method can also be improved
by a Kalman smoothing filtering [30]. For an accurate analysis of the joint kinematics, the
joints cannot be modelled as hinges or ball and sockets. In this case, the global optimization
method becomes less interesting compared with a local approach.
The marker placement can follow the ISB recommendations on the definition of joint
coordinate systems or be placed over joints. The segment orientation is based on anatomical
landmarks indirectly measured. In this chapter, we prefer to associate the CAST protocol
with the functional method for locating joint centre. The main advantage is to minimize soft
tissue artefact by putting technical markers far from joints. Furthermore, several studies
have highlighted that using more than three markers per segment is a good way for mini-
mizing noise and soft tissue artefacts [8, 23, 70]: the segment kinematics is then calculated
using minimization methods. However, these methods are not widely used in biomechanics
specially in clinical analysis where two markers per segment can be used while the third
one is a virtual marker and comes from assumption about joint centre locations.
The first section (2) of this chapter deals with the effect of using a redundant marker set
(n > 3) for locating joint centre. The functional method is presented (2.1.1) and then the
optimization method (2.1.2) for calculating rotation and translation from several markers
is detailed. Examples obtained for the thigh and the upper-limb kinematics highlight the
benefit of such an approach (2.2 and 2.3).
2.1 Method
2.1.1 Functional method for locating joint centre
The method for locating joint centres will be applied to various joints; the first example
concerns the hip joint centre. The positions of hip joint centre with respect to ( wrt) the
local reference frame of the pelvis BP and wrt the local reference frame of the thigh BT
are constant and denoted u and v, respectively (Figure 5). The matrix RP (RT ) is the
rotation between BP (BT ) and the global reference frame B0 and is calculated from the
coordinates of the six markers Pi (Ti ) using the optimization procedure described in the
following section. The vector pP (pT ) is the translation from B0 to BP (BT ). For each
frame f , we have the equality:

f u
RP −f RT = f pT − f pP . (3)
| {z } v | {z }
3×6 |{z} 3×1
6×1
The vectors v and u can be found with at least two different configurations. But more
configurations would be desirable because of noise. The vectors u and v are therefore
estimated in a least square sense:

u −1 T
= BT B B C, (4)
v
Figure 5: Markers positions and vectors used in the SCoRE method. Vector u is the position
of hip joint centre wrt BP and vector v is the position of hip joint centre wrt BT .
with
1 
RP −1 RT
 .. ,
B =  ... .  (5)
nR n
P − RT
| {z }
3n×6
and
1 
pT − 1 pP
 .. 
C = . . (6)
np − n p
T P
| {z }
3n×1
For an accurate estimation of the joint centre location, this equation highlights the impact
of the rotation matrix (RP , RT ) and translation vectors pT , pP for each segment. The
purpose of the following section is to propose an optimization method for determining
these parameters.
2.1.2 Rotation matrix optimization

Usually, rotation matrix of each segment is computed with three markers, that is the mini-
mum. However, in order to limit the uncertainty associated to the motion analysis system
and to minimize soft tissue artefact, a few authors have presented optimization methods
for calculating rotation matrices with more than three markers [21, 22, 66, 70]. Another
method coming from mathematics [14] has the advantage to not using square root or in-
verse operation. This method is detailed in the following paragraphs.
M3
M4
ai M2 S1 (t1 )
M1
B1
rM1
z0 pi M2
M4
M3
bi
B0 Bn S1 (tn )
M1
qi
O
y0
x0
Figure 6: Vector definition for one segment-embedded reference at two times ( t1 and tn ).
All the notations refer to Figure 6. The first step is to calculate the rotation of segment
S1 at time t1 with respect to B0 (t01 R). The rotation matrix t01 R between B1 and B0 is:
t1

0 R= x y z , (7)
where
rM2 (t1 ) − rM1 (t1 )
x= , (8)
krM2 (t1 ) − rM1 (t1 )k
rM3 (t1 ) − rM1 (t1 )

z = x× , (9)
krM3 (t1 ) − rM1 (t1 )k
and
y = z × x. (10)
In the second step, the rotation matrix R is the rotation of the segment S1 from tn to t1 :
R = ttn1 R. The rotation matrix of the segment is such that pi = Rqi + v and ai = Rbi . R
can be found by maximizing the following function [23]:
" n #
1X
tr bi ai R = tr C T R ,
T
(11)
n
i=1
(tr represents the trace operator) with

n
1X
C= bi aTi , (12)
n
i=1
n
1X
ai = rMi − rMi (13)
n
i=1
and
n
1X
bi = rMi − bMi . (14)
n
i=1
In the present chapter, a procedure different to that proposed in [23] is implemented for
calculating R without use of the square root or inverse operations based on the work of [14].
Firstly, the symmetric part of C and the quadrivector w associated to the skew symmetric
part of C are computed as
1
S = (C + C T ) (15)
2
and
j(w) = C − C T , (16)
where j(w) is the skew-symmetric mapping defined by: j(w)v = w × v. To the 3 × 3

matrix C is associated the 4 × 4 matrix Q,

2S − [tr(S)] 13 w
Q= . (17)
wT tr(S)
Any rotation can be expressed by the formula:

R = m20 − m.m 13 + 2m.mT + 2m0 [j (m)] , (18)
with m.m + m20 = 1 and m0 > 0 (13 ) is a 3 × 3 identity matrix. To the rotation R is
associated the quadrivector Y [61]:

m
Y= . (19)
m0
With the association of Qand Y to C and R the expression for tr(C T R) in Equation 11
becomes:

tr C T R = Y.QY. (20)
The problem formulated in Equation 11 is now to maximize:

Y.QY
φ(Y) = (21)
Y.Y
The conjugate gradient iterative algorithm is then applied [15]:
• Initialise Y0, W0 = G0 = (gradφ) (Y0 )
• While Gi >
- Search for µi maximizing φ (Yi + µWi )

- Compute Yi+1 = Yi + µi Wi
- Compute Gi+1 = (gradφ) (Yi+1 )
- Compute λi+1 = Gi+1 .(G i+1 −Gi )
Gi .Gi
- Determine the new search direction Wi+1 = Gi+1 + λi+1Wi
• end
In this algorithm, G and µ are calculated by the formulae:
2
G= [QY − φ(Y)Y] (22)
Y.Y
and
√
−β − ∆
µ= . (23)
2α
In these formulae,
α = (Y.W)(QW.W) − (QY.W)(W.W), (24)
β = (Y.Y)(QW.W) − (Y.QY)(W.W), (25)

and

(Y.W)2
∆ = 1− β 2 + 4(Y.Y)(W.W)
(Y.Y)(W.W)
2
(Y.W) (Y.Y)(QW.W) + (W.W)(QY.Y)
QY.W − . (26)
(Y.Y)(W.W) 2
The rotation matrix ttn1 R is then calculated from the solution (the quadrivector Y) of the
gradient algorithm with Equation 18. In a third step, at each instant of time the optimized
rotation matrix t0n R between Bn and B0 is:
tn
0 R = t01 R tn
t1 R (27)
In current biomechanical analysis, the rotation matrix is calculated with three markers
using Equations 7 to 10 at each instant of time and the translation vector is the position in
the global reference frame of coordinates of one of these three markers:
t = rMi . (28)
Whereas in the optimal methods, the translation vector is defined as the mean vector of all
the markers fixed on the segment:
n
1X
t= rM i . (29)
n
i=1
To summarize, the conjugate gradient iterative algorithm proposed here calculates the
rotation between two orientations of the segment without square root or inverse operations.
The orientation of the segment is then recalled in the global reference frame with Equation
(27).
2.2 Effects of the rotation matrix on the kinematics

To illustrate the benefit for calculating the rotation matrix from a redundant marker set, a
movement was simulated with a noise added to a reference kinematics. The artificial noise
included random and continuous components, it was designed to reproduce skin movement
artefact plus measurement errors (see [7] for details). The rotation matrix was then calcu-
lated from the simulated data (reference) and from noisy data with a number of markers
ranging from three to eight. In a paper dealing with the determination of the rotation ma-
trix, Challis [23] proposed to assess the accuracy in the kinematics estimation parameters
by calculating the mean absolute relative difference E between the data from the noisy
kinematics and those obtained from the reference kinematics. The average of the absolute
difference was calculated by the following formula:
n
1 X V Ci − V Ei

E= (30)
n V Ci
i=1
where n is the number of data sample, V Ci is the real value (obtained from the reference
kinematic) and V Ei is the estimated value obtained from the noisy data.
The mean absolute relative difference was calculated on three different sets of kinemat-
ical entities: the nine coefficients of the rotation matrix ( R), the angular velocity vector (ω)
and the nine coefficients of the second time derivative of rotation matrix ( R̈). The effect of
an increased number of markers for calculating the rotation matrix is not very pronounced
when regarding the nine coefficients of the rotation matrix (Figure 7): the mean absolute
relative difference is 0.028 for three markers and drops to 0.011 for seven markers. These
results as those of Challis [23] show that the accuracy increases with the number of markers
and the change in accuracy was most rapid when the number of markers increased from
three to four. We can also observe that the number of markers is not always correlated with
an improvement of the accuracy because an additional marker can give more noise than
extra information. For example from five to six markers, the error drives from 0.0139 to
0.0146. However, this trend disappears with the time derivation. For the angular velocity
and for the second time derivative of the rotation matrix, an increased number of markers is
always related to a better accuracy.
This simulation study was continued with an experimental approach [46]. A set of 26
markers was methodically fixed on the left thigh of a male participant who was asked to
perform several walking trials. The hip joint kinematics were calculated with various com-
binations of three to six markers while a reference kinematics was calculated with all the 26
markers. The marker locations and the effect of using more than three markers were evalu-
ated for flexion-extension, abduction-adduction and mediolateral rotation angles (Figure 8).
From three to six markers the precision increased: from 3.20to 2.20for flexion-extension an-
gle, from 3.18to 2.07for abduction-adduction angle and from 7.33to 3.73for mediolateral
hip rotation angle (reference angles were obtained with the 26 markers). Statistical differ-
ences (ANOVA, p<0.05) were observed between three and four markers, between three
and five markers and between three and six markers for all the axes of rotation.
These results highlight the benefit of using more than three markers for calculating
the rotation matrix. It has a direct impact on the nine coefficients of the rotation matrix.
When hip joint angles are calculated, these results show that the precision is increased
0.03
0.02
R
0.01
3 4 5 6 7 8
0.04
ω 0.02
0
3 4 5 6 7 8
4
R̈ 2
0
3 4 5 6 7 8
Number of Markers
Figure 7: Mean absolute relative difference for the rotation matrix ( R), the angular velocity
(ω) and the second time derivative of the rotation matrix ( R̈) according to the number of
markers.
∗
z }| {
RMS [degres]
z }| {
∗ ∗
z }| { z }| {
z }| { z }| {
Number of markers
Figure 8: Mean RMS differences of the three angles (flexion-extension, abduction-

adduction and mediolateral rotation from left to right) according to the number of markers
involved.
with the number of markers. Statistically, the RMS difference is equal for four, five and
six markers. For describing the segment kinematics or for locating joint centre, it appears
that a redundant marker set composed of four or more markers improves significantly the
accuracy and avoid to apply filters on the raw coordinates. The next step will be to define
the precise locations of this markers set.
2.3 Results on joint centre location

For assessing the effects of the marker placement and the number of markers involved in
the rototranslation parameters, joint centres were located using two approaches: the first
approach has consisted in locating the joint centre using the most common way i.e. the
rotation matrices of adjacent segments are calculated from three markers (Equations 7 to
10) and the translation vector is the coordinates of one marker. The second approach has
located the joint centre with the optimal method: the rotation matrix is calculated with
six markers and the translation vector is the centroid of the six markers. Two studies are
presented here [8, 57]. The first one is focused on the shoulder and elbow joint centre
locations with all the combinations of three markers for both proximal and distal segments.
The second study applies the optimal method (rotation matrix and translation vector) for
locating the hip joint centre.
In the first study [57], the authors evidenced that the SCoRE method (as all the func-
tional methods) located differently the joint centre according to the selected three markers
(for both proximal and distal segments) and their order. Results concern the reconstruc-
tion of the joint centres during a fast movement with impact, i.e. a punch. In a first step,
multiple joint centres were estimated from the set-up movement data for all permutations
of three from six markers in both proximal and distal segments. Groups of three markers
that were used to define a local reference frame were called triads. Then, all these solutions
were reconstructed in a global frame during the movement of interest and these locations
were fitted with a 95% sphere of precision. This procedure was performed for the shoulder
and elbow joint centres. Most of the time using three markers ( Mi,j,k ), the local systems
of co-ordinates are calculated by Equations (7) to (10). Figure 9 is a two-dimensional ex-
ample to illustrate how the noise associated to each marker can affect the local system of
co-ordinates differently according to the marker order. Since the noise is three-dimensional
and marker-specific, diffences in the joint centre location is up to 44 mm (shoulder) or
56 mm (elbow) for all the permutations of a triad. For some triads the maximum radius is
lower than 15 mm for both joints but never nought (Figure 10). Although solidification [26]
has reduced random error it has dispersed the joint centre locations further apart, especially
along one axis of the elbow. The actual information has also been reduced.
Using only three markers, these results would indicate that determining the joint centre
with an accuracy of greater than approximately 20 mm is unlikely. This result confirms
our affirmation in the introduction, the quality and accuracy of the kinematics is of highest
importance.
It is also pertinent to note that, although the theoretical literature regarding functional
methods relates to hinges or ball and sockets joints, the elbow, knee or ankle are neither
truly hinge nor ball and socket joints. However, as presented in [57], the elbow is not
ideally suited to determine joint centre location. But the SCoRE method can be applied to
M3
Marker reference
JC Joint centre
Small error in M2
M1 M2 M2
NO EFFECT
M3 Order M1 , M2 , M3
y
x = M2 − M1
JC y0 y0 = M3 − M1
z = x ∧ y0
y =z∧x
x
M1 M2 M2
EFFECT OF ERROR IN ORIGIN

y0 y x = M1 − M2
JC y0 = M3 − M2
z = x ∧ y0
y =z∧x
M1 x M2 M2
EFFECT OF ERROR IN ORIGIN + ORIENTATION

x = M3 − M2
JC x
y0 = M1 − M2
z = x ∧ y0
y =z∧x
M1 y0 M2 M2
Figure 9: Effect of the noise associated to one marker (M2) on the joint centre position (JC)
in a two-dimensional case. When the marker order is M1 , M2 , M3 , the error on M2 has no
influence on the joint centre location. When the marker order is M2, M1 , M3 , the error on
M2 affects the location of the joint centre but does not affect the segment reference frame.
When the marker order is M2 , M3 , M1, the error on M2 affects both the rotation and the
translation.
Figure 10: Lateral views of the upper-limb and the 14,400 locations of shoulder and elbow
joint centres for the 1 st frame (black lines and dots) and the 220 th frame (grey lines and
dots) of the punch [57].
the elbow joint given that the combined movement from each of the three articulations was
not uniaxial.
The second experimental application is related to the hip joint centre location and has
assessed the effect of using more than three markers for calculating the rotation matrix.
For that, six reflective markers were fixed on the thigh and six on the pelvis (Figure 5).
The subject was asked to perform three slow star-arc movements [16] over a large range of
motion. Hip joint centre was located using the SCoRE method with all the combination of
three markers on the distal and proximal segment and using one marker for the translation
vector (Classical method). From the same data, hip joint centre was located using all the
six markers and the centroid vector for the translation vector ( Optimal method). As the
order of the markers used to calculate the rotation matrix affects the orientation of the
segment-embedded reference frame [57], for all the permutation of markers in the proximal
and distal segment, a joint centre location was calculated. Like in the previous study, a
sphere of precision was calculated in static and dynamic conditions (during walking) for
the Classical and Optimal methods.
In static, the locations of the hip joint centres of the spheres varied slightly with the
method: 3.7 mm (Table 1).
However the radii of the static spheres ranged from 4.0 mm to 43.5 mm and during the
walking trial, the radii of the dynamic spheres increased to 45.6 mm for Classical method
and to 29.9 mm for Optimal method (Table 1). The estimation of an optimal joint centre
location is affected by the number of markers and their placement. However the main
uncertainty comes from the reconstruction of this optimal location in the motion of interest
where only the redundancy of the marker set can provide an accurate location.
2.4 Conclusion
All these results highlight the effects of the numbers of markers involved in the kinematics
calculation, i.e. joint centre locations and angles. Increasing the number of markers is
important for an accurate estimation of the rototranslation matrix.
The rotation matrix of each segment has an impact on the angles determination and on
Table 1: Location and precision [mm] of the hip joint centre reconstruction for the static
and walking trial using classical and optimal methods [8].
Classical method Optimal method
3*Static location x 115.1 113.0
y 300.5 303.4
z 897.5 898.5
2*Static radius min 40.0 4.0

max 43.5 4.2
2*Dynamic radius min 31.9 15.5

max 45.6 29.9
the joint centre locations when using SCoRE method. Our results show the inter-marker
variability on the kinematics due to soft tissue artefact that cannot be directly measured
without invasive approaches. Soft tissue artefact is defined as markers movement relative
to the underlying bone. Currently only invasive devices and fluoroscopic techniques can
achieve direct measurement of bony segment movement [37]. Error introduced in the joint
centre reconstruction came mainly from the lack of redundancy during the reconstruction
while the optimal joint centre is calculated for numerous frames. Since joint centre cannot
be determined from the motion of interest [54], subjects should perform extra set-up trials
which have different skin movement artefacts than those during the movement of interest.
This is another source of error; the noise has to be minimized in both movements.
The accuracy of the joint centre location is not only due to the number of markers but
also due to the movement characteristics as shown in the results of the first study [57]. Since
the joint centre is obtained by a pseudo-inverse procedure, the redundancy of the movement
is important while the noise remains small. Once the joint centre locations have been de-
termined, a method of reconstructing them during the movement of interest is normally
required. In this second step, the accuracy is not improved by the movement redundancy
because the joint centre is reconstructed in each single frame. The studies of Begon et al.
[8] and Roosen et al. [57] assess this difficult and blank part of the biomechanical anal-
ysis. They recall important principles of human biomechanics analysed using rigid body
mechanics. By in vivo examples they have shown the limits of the solid mechanics for
biomechanics and have proposed methods for improving the accuracy of the kinematics.
Since the human body segments are not rigid, three markers are not enough to calculate
accurately the kinematics. The classical approach for calculating the rototranslation matrix
using three markers (Equations 7 to 10) holds concurrently error on the origin and on the
reference axes. According to the marker placement on the segment and to their order (i.e.
ijk, ikj, jki, jik, kij, kji for markers i, j, k), the joint centre location and the joint kine-
matics will differ. The equality between all the permutations occurs only in one case: the
body is perfectly rigid and there is no noise in the marker coordinates measurement. This
case is impossible in biomechanics. The hypotheses assessed in [8, 57] are not exhaustive
and will be completed in the next future.
3 Body segment parameters

Are the body segment inertial parameters essential in biomechanics? As said in a recent
paper [55] the influence of body segment inertial parameters for solving the problem of
inverse dynamics is controversial. Some papers reported low importance ( < 0.06 Nm.kg-1 )
of body segment inertial parameters uncertainties [24, 36, 53] while others demonstrated
the opposite (> 12%) [1, 40, 60]. The difference between these two findings comes mainly
from the movements that each group studied. During walking, low accelerations of lower-
limbs and high importance of the reaction force during step minimize effects due to body
segment inertial parameters on net moments at the hip, knee or ankle joints. While during
other movements [1, 40], specially without contact forces, body segment inertial parameters
show non negligible effects on net moments.
Whatever is our opinion and experience on this question, body segment inertial pa-
rameters should be estimated for any biomechanical analysis. The researcher can choose
between several models or methods: geometric models, proportional models derived from
cadavers’ studies, in vivo scanning models or identification. Each approach has its advan-
tages and drawbacks (see [13] for an overview). We propose to present and develop the
latter method – the identification – for the determination of personalized body segment in-
ertial parameters. It consists in solving a redundant system for the inertial parameters by a
numerical computation. This method has been already used by [65] but with some limita-
tions and assumptions (two-dimensional model, symmetric limbs). Originally identification
was developed for estimating the segment parameters of robots. Indeed, for robot manip-
ulators, these parameters are not always precisely known [6, 56] and various algorithms of
identification exist to determine them [2, 3, 6, 12, 25, 56]. In the classical identification
approaches [2, 6], the parameters are estimated from motion data and actuator torques or
forces, both measured by internal measurement devices. For human applications, this ap-
proach is impossible because net joints forces and moments cannot be measured; they are
calculated from kinematics and contact forces using an inverse dynamics procedure. An
alternative approach is the external identification. In this method, the external model relates
the motion of the multibody system to the reaction forces and torques on a force platform.
This technique was initiated by [56] for identifying the segment inertial properties of a three
segment robot (PUMA 562).
The method proposed by [56] is based on the recursive Newton-Euler algorithm in
which the inertial parameters of each segment are represented by a ten-dimensional vector
solved by a least-squared method (pseudo-inverse). The ten inertia parameters of n seg-
ments are contained in the x vector (which is a 10 n dimensional vector) and the following
inverse problem is solved for x:
Ax = b, (31)
where matrix A contained the kinematics and vector b groups the external forces and
torques together. This method was applied to the human body with degrees of freedom
ranging from 6 to 25 [25] but the results were unsatisfactory on account of the high number
of degrees of freedom. Furthermore, when the inertia parameters are gathered in a vector,
the properties of the inertia matrix (i.e., symmetric and positive definite) are lost. That is
why we prefer a matrix form of the motion equations that keeps the properties of the inertia
matrix and use them as constraints during the resolution of the inverse problem.
Some authors introduced the matrix formulations for dynamical analysis. [41] [43] and
proposed 4 × 4 and 6 × 6 matrices, respectively. Fayet et al. [34, 35] have introduced the
global inertia tensor for an iterative formulation of the equations of motion in a 4 × 4 matrix
form. However all these methods are recursive formalisms. Since they are applied succes-
sively to the different segments, they cannot be used for human applications where the net
joint forces and moments are unknown. In answer to this weakness, we propose an original
expression of the Newton-Euler equation of motion where all the inertial parameters of a
multibody system composed of n segments are grouped together in a single symmetric and
positive definite matrix of dimension (3n + 1) × (3n + 1) [5].
The external identification appears favourable to the determination of the human body
segment inertial parameters. The high number of the human body segments could be ar-
tificially reduced by soliciting a single joint. In the following example, the human body
is modeled as a two-segment system: the right upper-limb and the remainder of the body.
The purpose of this section is to detail the matrix formulation of the equation of motion
for identifying the inertia parameters of the upper-limb. The following section is organized
in three parts: firstly the matrix formulation of the motion is detailed (3.1) and then the
subsections 3.2 and 3.3 present the identification process and the experimental protocol.
3.1 An original Newton-Euler formalism for identification

Let S1 and S2 be two rigid bodies of mass m1 and m2 coupled each other around the joint
centre A which is common to both bodies (Figure 11). These bodies move in a Galilean
Figure 11: Schema
reference frame with reference frame R0 (O, x0, y0, z0 ). The mass centre of these bodies
are denoted G1 and G2 and are expressed with respect to the joint centre by the vectors
−→
ci (i = 1, 2). The position of the joint centre A is described by the vector: rA = OA
(note: the different reference frames are expressed by a superscript in parentheses). The
T
relation between the global coordinates of a vector c(0) = cx0 cy0 cz0 in the abso-
(i)
T
lute reference frame R0 and its local coordinates c = cxi cyi czi expressed in
the body-fixed reference frame, is described by the rotation matrix R : c = R(i)c(i)
(i) (0)
(Figure 11).
The position of any point M of S2 is expressed with respect to A by the vector b2 =
−−→
AM . According to Vallée et al. [63], the position of M with respect to the reference frame
R0 can be written with the rotation matrix R(2):
−−→ −→ −−→
OM = OA + AM
(0) (2)
= rA + R(2)b2 . (32)
With this notation, the acceleration of M noted ΓM is simply:

(0) (2)
ΓM = r̈A + R̈(2)b2 . (33)
So, the dynamic resultant of S2 is:

Z
(0) (2)
χ2 = (ΓM ) dm = m2 r̈A + R̈(2)c2 , (34)
S2
R
where m2 c2 = S2 (b2 ) dm defines the mass centre. With the same notation, we can
express the dynamic resultant of S1 :
Z
(0) (1)
χ1 = (ΓM ) dm = m1 r̈A + R̈(1)c1 . (35)
S1
In the inertial frame R0, the Newton second law is applied successively to S1 and S2:

(0) (0) (0) (0) (2) (2)
F1 = χ2 + m2g = m2 r̈A + g + R̈ c2 (36)
(0) (0)
F0 = χ1 + m1g(0) + F1

(0) (1) (2)
= (m1 + m2 ) r̈A + g(0) + m1 R̈(1)c1 + m2R̈(2)c2 , (37)
where g(0) is the gravity acceleration: g(0) = [0 0 9.81]T m.s−2 and Fi is the reaction
force between Si and Si+1 . Equation 37 can be written in matrix form:
 
(2)
h i m2c2
(0)   (0)
R̈(2) R̈(1) r̈A + g(0)  m1c(1)
1
 = F0 . (38)
m1 + m2
For the rotation, we choose to express the moment by a 3 × 3 skew symmetric matrix,
using the tensorial product of two vectors defined by:
a(⊗)b = abT . (39)

So the dynamic moment of S2 about A (noted δA (S2)) is:

Z h i
(2)
δA (S2) = skew R(2)b2 ∧ ΓM dm
Z hS2 i
(2) (2)
= ΓM ⊗ R(2)b2 − R(2)b2 ⊗ ΓM dm. (40)
S2
In this equation, skew is the operator which associates to any trivector a 3 × 3 skew-
symmetric matrix:
   
X 0 −Z Y
skew  Y  =  Z 0 −X  . (41)
Z −Y X 0
Using Equation (33), Equation. (40) becomes after simplification:

Z h i
(0) (2) (2) (2)
δA (S2) = r̈A ⊗ m2 R(2)c2 + R̈(2)b2 ⊗ R(2) b2 dm −
S2
Z h i
(2) (2) (0) (2) (2)
m2 R c2 ⊗ r̈A − R(2)b2 ⊗ R̈(2) b2 dm. (42)
S2
Thanks to the notion of transposition:

(0) (2) (0) (2)
r̈A ⊗ m2 R(2)c2 = r̈A ⊗ m2 c2 R(2)T , (43)
we can rewrite Equation (42) as follows:

Z
(0) (2) (2)T (2) (2) (2)
δA (S2 ) = r̈A ⊗ m2 c2 R + R̈ dm R(2)T −
b2 ⊗ b2
S
Z 2
(2) (2) (0) (2) (2) (2)
R m2c2 ⊗ r̈A − R b2 ⊗ b2 dm R̈(2)T . (44)
S2
(2)
The integral in brackets is the definition of the Poinsot inertia matrix and is noted K2 :
Z
(2) (2) (2)
K2 = b2 ⊗ b2 dm, (45)
S2
and represents a form of the inertia tensor of the body at A2 . K2 is related to the classical
definition of the inertia tensor J2 by the relation:
1
K2 = (trJ2 ) I − J2 , (46)
2
where I is the identity matrix (see [63] for details). The dynamic principle is written as:
h i
(0) (2) (2)
MA (S2) = r̈A + g(0) ⊗ m2 c2 R(2)T + R̈(2)K2 R(2)T −
h i
(2) (0) (2)
R(2) (m2 c2 ⊗ r̈A + g(0) − R(2)K2 R̈(2)T , (47)
where MA (S2) is the moment exerted by S1 on S2 calculated in A and expressed in the

reference frame R0. MA (S2 ) is the 3×3 skew-symmetric form of the usual moment vector
MA (S2):
MA (S2 ) = skew (MA (S2)) , (48)
that is to say:
 
MA (S2)x
MA (S2) =  MA (S2)y  (49)
MA (S2)z
 
0 −MA (S2)z MA (S2 )y
MA (S2) =  MA (S2)z 0 −MA (S2 )x  . (50)
−MA (S2)y MA (S2 )x 0
With the same notations, the dynamic moment of S1 about A reads:

(0) (1) (1)
δA (S1) = r̈A ⊗ m1c1 R(1)T + R̈(1)K1 R(1)T −

(1) (0) (1)
R(1) m1 c1 ⊗ r̈A − R(1)K1 R̈(1)T . (51)
Using classical inverse dynamics, the moment at the point A due to S0 is:
h i
(0) (1) (1)
MA (S0) = r̈A + g(0) ⊗ m1c1 R(1)T + R̈(1)K1 R(1)T −
h i
(0) (1)
R(1) m1 c(1) ⊗ r̈A + g(0) − R(1)K1 R̈(1)T + MA (S2 ) (52)
The equations of translation (1) and rotation (2) can be summarized in matrix form by
introducing the 4 × 7 matrices A and B:
" #
(0)
R̈(2) R̈(1) r̈A + g(0)
A = , (53)
0T 0T 0
(2)
R R(1) 0
B = . (54)
0T 0T 1
Let us summarize the inertia parameters of bodies S1 and S2 in the 7 × 7 matrix H:
 (2) (2)

K2 [0] m2 c2
 (1) (1) 
H= [0] K1 m1 c1 . (55)
(2)T (1)T
m2 c2 m1 c1 m1 + m2
The equations of motion are then written:
" #
(0)
MA (S0) F0
AHB T − BHAT = C = (0)T . (56)
−F0 0
Matrices A, B and C are time dependents. Matrix H is constant because all of this
parameters are expressed in the body fixed reference frame. So the translational and rota-
tional motions equations are written in matrix form and the inertia parameters of S1 and S2
appear therein linearly. The matrix C is easily available with a force platform and matrices
A and B are determine with a motion analysis system. So matrix H could be found by a
minimization procedure.
3.2 Conjugate gradient algorithm for identifying a symmetric and positive

definite matrix
The matrix C is a 4 × 4 skew-symmetric matrix, matrix C has six independent components.
The matrix H is a 7 × 7 symmetric and positive definite matrix, so matrix H has 28 inde-
pendent components. Therefore Equation (56) provides six conditions for 28 unknown, but
these six conditions hold at each time. This equation must be sampled in at least five dif-
ferent configurations; more configurations would be desirable because of noise. We make
n experiments where ti is the duration of the ith experiment. Using a least squared method,
we define the function J(H):
n Z
1X ti
Ai (t)HBi (t)T − Bi (t)HAi(t)T − Ci (t) 2 dt
J(H) = (57)
4 0
i=1
This function J is minimized by the following conjugate gradient algorithm:
1. Initialize H with H0
G0 = ∇J(H0 )
D0 = −G0 ; k = 0
2. iteration k
tr(GT
k Dk )
µk = PR
tr[DkT ( BADk BA−BB T Dk AT A)dt]
Hk+1 = Hk + µk Dk
3. if kGk+1 k < break, else
Dk+1 = −Gk+1 + λk Dk
tr[GT
k+1 (Gk+1 −Gk )]
with λk = tr(GT
k Gk )
k = k + 1 et return in 2.
Remark: In µk , indices, time and duration are omitted.

In this algorithm, the gradient (∇) of the function J is the matrix calculated as follows:
n Z
X ti
BiT (t)Ai (t)HBiT (t)Ai (t) + ATi (t)Bi (t)H T ATi (t)Bi (t)
∇J(H) = − dt +
2
i=1 0
Xn Z ti T
Bi (t)Bi (t)HATi (t)Ai (t) + ATi (t)Ai (t)H T BiT (t)Bi (t)
dt +
2
i=1 0
Xn Z ti T
Bi (t)Ci (t)Ai (t) + ATi (t)CiT (t)Bi (t)
dt (58)
0 2
i=1
3.3 Application on the upper-limb

Three men participated in this study. They were instructed to perform fifteen cycles of a
complex movement of the right upper-limb composed of a flexion, horizontal abduction,
adduction and mediolateral rotation. This movement was chosen because the three degrees
of freedom of the glenohumeral joint were excited. This condition is necessary for identify-
ing all the inertia parameters of the upper-limb. Subjects were sat on a chair and they were
asked to keep their trunk and legs as fixed as possible and to move only the right upper-limb.
A Saga3RT motion analysis system (Biogesta, Valenciennes–France) was used to col-
lect kinematical data with six infrared cameras (50 Hz) inside a calibrated volume of 2 m
× 1 m × 1 m. Force platform data (Kistler 5233A) were synchronized. Ten markers were
fixed on the body: four on the right upper-limb, three on the thorax and three on a rigid tri-
pod fixed to the acromion. The tripod was used to identify the scapula motion for locating
the glenohumeral joint centre using the SCoRE method. Since only the right upper-limb
(2)
was moving, only ten inertia parameters were excited: K2 , m2c2 and m1 + m2 (see Equa-
tion (55)). As observed in the Figure 12, the solution is obtained in ten iterations because
the conjugate gradient is a direct method. After the tenth iterations, the cost function is
Figure 12: Evolution of the cost function during the iterations.
constant and the solution is obtained.

The inertia matrix identified is the Poinsot inertia matrix:
Z
(2) (2) (2)T
K2 = b 2 b2 dm. (59)
S2
(2)
The usual inertia matrix is obtained from K2 by the formula [63]:

(2) (2) (2)
J2 = trace K2 13 − K2 (60)
Table 2 presents the inertia matrix and the product of the mass by the centre of mass
position (m2 c2 of the upper-limb obtained with the identification method. These parameters
are compared with upper-arm inertial parameters estimated with the proportional model of
Zatsiorsky modified by de Leva [31].
For two subjects (S1 and S3), the inertia moment along the longitudinal axis seems to
be overestimated compared to the classical body segment inertial parameters. This result
Table 2: Inertia matrix (kg.m2) and first order moment (m2 c2 in kg.m) of the upper-limb
for the three subjects obtained with the identification method and with the proportional
model of Zatsiorsky modified by de Leva [31].
de Leva Identification
   
0.454 0 0 0.476 0.001 −0.018
S1 (1.80 m, 70 kg)  0 0.008 0   0.001 0.040 0.035 
0 0 0.450 −0.018 0.035 0.486
T T
0 1.0734 0 0.1123 1.2210 −0.0485
   
0.495 0 0 0.504 0.021 −0.001
S2 (1.78 m, 78 kg)  0 0.009 0   0.021 0.006 −0.025
0 0 0.490 −0.001 −0.025 0.503
T T
0 1.1954 0 −0.0429 0.7603 0.0201
   
0.476 0 0 0.512 0.005 0.006
S3 (1.78 m, 75 kg)  0 0.008 0  0.005 0.033 0.058
0 0 0.472 0.006 0.058 0.528
T T
0 1.1293 0 0.1200 1.0285 0.0787
could be caused by a small elbow flexion. Unfortunately no marker was fixed to the forearm
in order to verify this hypothesis. Identified inertia and first order moment (m2c2 ) values
are realistic values. The main advantage of the identification method is to propose products
of inertia and a centre of mass not aligned with the longitudinal axis of the segment. This
signified that segment are not homogeneous and not symmetric whereas they are mainly
represented in most of the geometric or proportional models. The drawback of this method
is to not determine the mass of the segments. Since the centre of mass position and the mass
do not appear independently in the Equation 55, only the first order moment is identified
and it is not possible to dissociate the mass from the first order moment.
To conclude, the external identification is a recent approach for determining the hu-
man body segment inertial parameters. The experimental procedure requires only a motion
analysis system and a force platform. The identified values of the matrix of inertia for the
upper-limb are realistic values and close to the values estimated by Zatsiorsky’s anthropo-
metric model. The difficulty in interpreting results comes from the fact that there is no ”gold
standard” for the body segment inertial parameters. The values calculated from the model
of Zatsiorsky give an idea of the order of magnitude but do not represent actual values.
The presented identification method is a good way for the body segment inertial parame-
ters determination without any assumption about the density of the segment or the segment
shape.
4 Conclusion
This chapter has presented methods and algorithms for a more accurate estimation of human
kinematics and kinetics. The first section focused on the rotation matrix determination and
joint centre locations and the second section proposed a new approach for the estimation of
the body segment inertial parameters.
As all experimental sciences, biomechanics is dependent on the measurement tool evo-
lution. Nowadays, the high camera resolution and the high sampling rate allow new ap-
plications that have not been explored yet. Tracking ten markers on each segment during
walking is no longer a problem. Some specific activities (sports for example) may be more
difficult to analyse mainly because of marker self-occlusions. It was the purpose of the first
section to evidence that using more than three markers on each segment improves signif-
icantly the kinematic. This result was already presented in 1995 by Challis [23] but this
method is still not commonly used in human movement analysis. An increased number
of markers improves the kinematics, especially the time derivative data (angular velocity,
linear and angular acceleration) and improves the joint centre location.
Of course, our conclusion is not to put twenty-six markers on each segment as used in
[46] for every day analyses because it is too time consuming. This solution accounts on
the global or average segment kinematics but studies with a high number of markers will
help to determine the best methods and marker placements. The set of experimental studies
proposed in this chapter makes recommendation for an accurate human movement analysis:
• combination of flexion-extension, abduction-adduction and circumduction for esti-

mating an optimal joint centre location;
• SCoRE is an accurate algorithm for locating joint centre because it is not sensitive to
velocity and calculates an average from the optimal locations found in both proximal
and distal segments;
• use a redundant marker set combined with a technique of local optimization;
• definition of the origin of the local frame using the marker centroid.
The proposed external identification method is suitable for a personalized estimation

of body segment inertial parameters. Moreover, the procedure requires only materials cur-
rently used in biomechanical analysis, i.e. motion analysis system and force platform. The
procedure proposed in this chapter involved a matrix formulation of the equations of motion
for conserving the properties of the inertia matrix. This choice was relevant because all the
body segment inertial parameters are grouped together in a symmetric and positive definite
matrix. This help to restrict the solution (during the optimization procedure) to the cone of
symmetric and positive definite matrix. This study may be continued with a three-segment
model, for example by splitting the upper-limb in arm and forearm plus hand. This method
needs other specific movements before collecting the movement of interest. As the time
available with the subject (especially in clinical analysis) is sometimes short, this could be
a limitation of this method. The main advantage of this approach is that no assumption is
made about the density of the segment or about the shape of the segment. Indeed, the centre
of mass is not located in the longitudinal axis but small components on antero-posterior and
medio-lateral axes are evidenced.
5 Perspectives
This was not assessed in this chapter but the use of a redundant set of markers could be
an alternative to raw data filtering or smoothing. The choice of the method of derivation
or smoothing is never easy because the skin movement artefact has similar frequency to
the skeletal kinematics. Derivation increases the noise and the time derivative data must be
treated. Optimization method for calculating the rotation matrix is a method for decreasing
the skin movement artefact and also the measurement error of the stereophotogrammetric
system. In our results [8, 48, 57]) no treatment was applied to the raw data; future work will
evaluate if the time derivative data need smoothing when they are calculated through opti-
mized rotation matrix. Future work will use the method proposed in [46] in order to define
the best markers set placement for an accurate and reliable determination of the segment
kinematics from six markers for example. For that, inter-subjects and inter-movements
effects on the markers placement will be measured.
In parallel to the experimental part, the theory on the rotation matrix calculation has to
be further explored. In the method proposed, the rotation matrix between two orientations
of the segment analysed is calculated by optimization. But, since we need the rotation
matrix with respect to the global system of coordinates, we recalled this optimized rotation
matrix with the rotation matrix between the orientation of the segment at the first frame with
the global system of coordinates. An error in orientation associated with the first frame is
then introduced. Future studies may focus on techniques for reducing this systematic error
of orientation by using more key frames and calculating average axes.
As skin movement artefact affects differently each marker according to its location,
using a redundant set of markers decreases the global skin movement artefact and improve
the measurement of the underlying bone kinematics. This hypothesis can be achieved by
comparison with invasive or fluoroscopic techniques only. In most applications soft tissue
artefact has to be minimized because skeletal kinematics is needed (joint centre location,
anatomical landmarks recalling). The inertia parameters of the soft tissue in motion affect
the joint dynamics [52]. For specific applications like identification of the body segment
parameters [47, 49], movement of the segment including soft tissue motion contribution is
needed.
The matrix formulation of the equations of motion presented in this paper may be used
for a n segment model. However, as explained in [25], the accuracy of the identified pa-
rameters depends on the length of the kinematic chain. They showed that the accuracy of
the identified parameter strongly decreases above six to eight degrees of freedom in the
kinematic chain. It is clearly inappropriate to use this method for all the body segments
considering the high degrees of freedom. Because of the constraints associated to the ma-
trix form of the equations of motion, our method should be more robust and accurate for
multibody systems with several degrees of freedom. Although this method could not iden-
tify all the body segment parameters simultaneously, we expect that it should be used for
each limb. For most of the subject studied, constraints consisting of searching the same
body segment inertial properties for the right and left limb may be introduced. This could
improve the method in terms of reliability and easiness. The next evolution of this method
will focus on specific movements that must be performed for an accurate identification. As
the identification is sensitive to the input data, results of the optimization method are differ-
ent according to the movement realized. Body segment inertial parameters calculated with
the identification approach from several movements have to be compared with reference
parameters (medical imaging?) in order to select the optimal movements.
Finally, it might be interesting to combine the identification with functional methods to
determine with the same setup movements both the inertial parameters and the joint centre
locations.
References
[1] Andrews, JG; Mish, SP. Methods for investigating the sensitivity of joint resultants to
body segment parameter variations. Journal of Biomechanics, 1996 29(5), 651–654.
[2] Antonelli, G; Caccavale, F; Chiacchio, P. Systematic procedure for the identification

of dynamic parameters of robot manipulators. Robotica, 1999 17(4), 427–435.
[3] Arteaga, MA. Robot control and parameter estimation with only joint position mea-
surements. Automatica, 2003 39(1), 67–73.
[4] Arthurs, KL; Andrews, DM. Upper extremity soft and rigid tissue mass prediction
using segment anthropometric measures and dxa. Journal of Biomechanics, 2009
42(3), 389–394.
[5] Atchonouglo, EK; Vallée, C; Monnet, T; Fortuné, D. Identification of the ten inertia
parameters of a rigid body. Journal of Applied Mathematics and Mechanics , 2008
72(1), 22–25.
[6] Atkeson, CG; An, CH; Hollerbach, JM. Estimation of inertial parameters of manipu-
lator loads and links. International Journal of Robotics Research , 1986 5(3), 101–119.
[7] Begon, M; Monnet, T; Lacouture, P. Effects of movement for estimating the hip joint
centre. Gait & Posture, 2007 25 (3), 353–359.
[8] Begon, M; Roosen, A; Pain, MTG; Monnet, T. Effect of segment-embedded reference

frame definition on hip joint centre location during walking. Journal of biomechanics,
In Submission.
[9] Begon, M; Wieber, P-B; Yeadon, MR. Kinematics estimation of straddled movements
on high bar from a limited number of skin markers using a chain model. Journal of
Biomechanics, 2008 41(3), 581–586.
[10] Bell, A; Pedersen, D; Brand, R. A comparison of the accuracy of several hip center
location prediction methods. Journal of Biomechanics, 1990 23, 617–621.
[11] Benedetti, MG; Catani, F; Leardini, A; Pignotti, E; Giannini, S. Data management in

gait analysis for clinical applications. Clinical Biomechanics, 1998 13(3), 204–215.
[12] Bennis, F; Khalil, W; Gautier, M. Calculation of the base inertial parameters of closed-
loops robots. In Proceedings - IEEE International Conference on Robotics and Au-
tomation, 1992 1, 370–375.
[13] Bjornstrup, J. Estimation of human body segment parameters. 1996. Available from
Inter Tech-Report, http://www.cvmt.auc.dk/jorgen/PhD/.
[14] Bouby, C; Fortuné, D; Vallée, C; Pietraszkiewicz, W. Direct determination of the rota-

tion in the polar decomposition of the deformation gradient by maximizing a rayleigh
quotient. ZAMM Zeitschrift fur Angewandte Mathematik und Mechanik , 2005 85(3),
155–162.
[15] Brousse, P. Optimization in mechanics: problems and methods. North Holland Editor ,
Amsterdam, 1988.
[16] Camomilla, V; Cereatti, A; Vannozzi, G; Cappozzo, A. An optimized protocol for

hip joint centre determination using the functional method. Journal of Biomechanics,
2006 39(6), 1096–1106.
[17] Cappozzo, A. Gait analysis methodology. Human Movement Science, 1984 3, 27–50.
[18] Cappozzo, A. Three-dimensional analysis of human walking: experimental methods

and associated artefacts. Human Movement Science, 1991 10, 589–602.
[19] Cappozzo, A; Catani, F; Della Croce, U; Leardini, A. Position and orientation in space
of bones during movement: anatomical frame definition and determination. Clinical
Biomechanics, 1995 10, 171–178.
[20] Cappozzo, A; Catani, F; Leardini, F; Benedetti, MG; Della Croce, U. Position and ori-
entation in space of bones during movement: experimental artefacts. Clinical Biome-
chanics, 1996 11(2), 90–100.
[21] Carman, AB; Milburn, PD. Determining rigid body transformation parameters from
ill-conditioned spatial marker co-ordinates. Journal of Biomechanics, 2006 39(10),
1778–1786.
[22] Challis, JH. An examination of procedures for determining body segment attitude and
position from noisy biomechanical data. Medical Engineering & Physics, 1995 17(2),
83–90.
[23] Challis, JH. A procedure for determining rigid body transformation parameters. Jour-
nal of Biomechanics, 1995 28(6), 733–737.
[24] Challis, JH; Kerwin, DG. Quantification of the uncertainties in resultant joint mo-
ments computed in a dynamic activity. Journal of Sports Sciences, 1996 14(3), 219–
231.
[25] Chenut, X; Fisette, P; Samin, J-Cl. Recursive formalism with a minimal dynamic pa-
rameterization for the identification and simulation of multibody systems. Application
to the human body. Multibody System Dynamics, 2002 8(2), 117–140.
[26] Chèze, L; Fregly, BJ; Dimnet, J. A solidification procedure to facilitate kinematic

analyses based on video system data. Journal of Biomechanics, 1995 28, 879-884
[27] Croce, UD; Cappozzo, A; Kerrigan, C; Lucchetti, L. Bone position and orientation
errors: Pelvis and lower limb anatomical landmark identification reliability. Gait &
Posture, 1997 5, 156-157
[28] Davis, RB; Ounpuu, S; Tyburski, D; Gage, JR. A gait analysis data collection and
reduction technique. Human Movement Science, 1991 10(5), 575–587.
[29] Davis, RB; Ounpuu, S; De Luca, PA. Joint moments: Evaluation of ground reaction,
inertial and segmental weight effects. Gait & Posture, 1994 2(1), 58–58.
[30] De Groote, F; De Laet, T; Jonkers, I; De Schutter, J. Kalman smoothing improves
the estimation of joint kinematics and kinetics in marker-based human gait analysis.
Journal of Biomechanics, 2008 41(16), 3390–3398.
[31] de Leva, P. Adjustments to Zatsiorsky-Seluyanov’s segment inertia parameters. Jour-
nal of Biomechanics, 1996 29(9), 1223–1230.
[32] Ehrig, RM; Taylor, WR; Duda, GN; Heller, MO. A survey of formal methods for
determining the centre of rotation of ball joints. Journal of Biomechanics, 2006 39(5),
2798–2809.
[33] Ehrig, RM; Taylor, WR; Duda, GN; Heller, MO. A survey of formal methods for
determining functional joint axes. Journal of Biomechanics, 2007 40(10), 2150 –
2157.
[34] Fayet, M; Pfister, F. Analysis of multibody systems with indirect coordinates and
global inertia tensors. European Journal of Mechanics - A/Solids , 1994 13, 431–457.
[35] Fayet, M; Sandel, A; Maiffredy, L. Utilisation des tenseurs d’inertie globaux dans le
cas de systèmes à liaisons complexes et pseudo-paramètres. 17ème congrès français
de mécanique, 2005 Troyes.
[36] Ganley, KJ; Powers, CM. Anthropometric parameters in children: a comparison of
values obtained from dual energy x-ray absorptiometry and cadaver-based estimates.
Gait & Posture, 2004 19(2), 133–140.
[37] Gao, B; Zheng, N. Investigation of soft tissue movement during level walking: Trans-
lations and rotations of skin markers. Journal of Biomechanics, 2008 41(15), 3189–
3195.
[38] Kadaba, M; Ramakrishnan, H; Wooten, M; Gainey, J; Gorton, G; Cochran, G. Re-
peatability of kinematic, kinetic and electromyographic data in normal adult gait.
Journal of Orthopaedic Research , 1989 7, 849–860.
[39] Kadaba, MP; Ramakrishnan, HK; Wootten, ME. Measurement of lower extremity
kinematics during level walking. Journal of Orthopaedic Research , 1990 8(3), 383–
392.
[40] Kingma, I; Toussaint, H; De Looze, M; Dieen, J. Segment inertial parameter evalua-
tion in two anthropometric models by application of a dynamic linked segment model.
Journal of Biomechanics, 1996 29, 693–704.
[41] Kozlowski, K. Computational requirements for a discrete kalman filter in robot dy-
namics algorithms. Robotica, 1993 11, 27–36.
[42] Lee, MK; Le, NS; Fang, AC; Koh, MTH. Measurement of body segment parameters
using dual energy x-ray absorptiometry and three-dimensional geometry: An applica-
tion in gait analysis. Journal of Biomechanics, 2009 42(3), 217–222.
[43] Legnani, G; Casolo, F; Righettini, P; Zappa, B. A homogeneous matrix approach

to 3D kinematics and dynamics – I. theory. Mechanism and Machine Theory, 1996
31(5), 573–587.
[44] Lu, TW; O’Connor, J. Bone position estimation from skin marker co-ordinates using
global optimisation with joint constraints. Journal of Biomechanics, 1999 32(2), 129–
134.
[45] Meskers, CGM; van der Helm, FCT; Rozendaal, LA; Rozing, PM. In vivo estima-
tion of the glenohumeral joint rotation center from scapular bony landmarks by linear
regression. Journal of Biomechanics, 1997 31(1), 129–134.
[46] Monnet, T; Begon, M; Pain, MTG; Colloud, F. Iterative solidification procedure to

determine the optimal marker location. In 10th international symposium on the 3D
Analysis of Human Movement. Amsterdam 28-31 October , 2008.
[47] Monnet, T; Begon, M; Vallée, C. Identification des paramètres inertiels du corps

humain. 18ème congrès Frana̧is de Mécanique, Grenoble, 2007.
[48] Monnet, T; Desailly, E; Begon, M; Vallée, C; Lacouture, P. Comparison of the SCoRE

and HA methods for locating in vivo the glenohumeral joint centre. Journal of Biome-
chanics, 2007 40, 3487–3492.
[49] Monnet, T; Vallée, C; Lacouture, P. Identification of the inertia matrix of the upper
limb. 21th congress of the International Society of Biomechanics , Taipei, 2007.
[50] Neptune, R; Hull, M. Accuracy assessment of methods for determining hip movement
in seated cycling. Journal of Biomechanics, 1995 28, 423–437.
[51] Nikolova, GS; Toshev, YE. Estimation of male and female body segment parameters
of the bulgarian population using a 16-segmental mathematical model. Journal of
Biomechanics, 2007 40, 3700–3707.
[52] Pain, MTG; Challis, JH. The influence of soft tissue movement on ground reaction
forces, joint torques and joint reaction forces in drop landings. Journal of Biomechan-
ics, 2006 39(1), 119–124.
[53] Pearsall, DJ; Costigan, PA. The effect of segment parameter error on gait analysis
results. Gait & Posture, 1999 9(3), 173–183.
[54] Piazza, SJ; Erdemir, A; Okita, N; Cavanagh, PR. Assessment of the functional method
of hip joint center location subject to reduced range of hip motion. Journal of Biome-
chanics, 2004 37(3), 349–356.
[55] Rao, G; Amarantini, D; Berton, E; Favier, D. Influence of body segments’ parameters

estimation models on inverse dynamics solutions during gait. Journal of Biomechan-
ics, 2006 39(8), 1531–1536.
[56] Raucent, B; Campion, B; Bastin, G; Samin, J-C; Willems, PY. Identification of the
barycentric parameters of robot manipulators from external measurements. Automat-
ica, 1992 28(5), 1011–1016.
[57] Roosen, A; Pain, MTG; Begon, M. Limitations of functionally determined joint cen-
tres for analysis of athletic human movement: A case study of the upper limb. Journal
of Applied Biomechanics, In Press.
[58] Rouhani, H; Favre, J; Crevoisier, X; Jolles, BM; Aminian, K. Validation of multi-

segment models for foot and ankle complex. Journal of Biomechanics, 2008 41(Sup-
plement 1), S168–S168.
[59] Roux, E; Bouilland, S; Godillon-Maquinghen, A-P; Bouttens, D. Evaluation of the

global optimisation method within the upper limb kinematics analysis. Journal of
Biomechanics, 2002 35, 1279–1283.
[60] Silva, MPT; Ambrosio, JAC. Sensitivity of the results produced by the inverse dy-
namic analysis of a human stride to perturbed input data. Gait & Posture, 2004 19(1),
35–49.
[61] Souriau, J-M. Géométrie et relativité. Editions Hermann, Paris, 1964.
[62] Stagni, R; Fantozzi, S; Cappello, A; Leardini, A. Quantification of soft tissue artefact

in motion analysis by combining 3d fluoroscopy and stereophotogrammetry: a study
on two subjects. Clinical Biomechanics, 2005 20(3), 320 – 329.
[63] Vallée, C; Hamdouni, A; Isnard, F; Fortuné, D. The equations of motion of a rigid

body without parametrization of rotations. Journal of Applied Mathematics and Me-
chanics, 1999 63(1), 25–30.
[64] Vaughan, CL; Davis, BL; O’Connor, JC. Dynamics of Human Gait. Human Kinetics
Publishers, 1992.
[65] Vaughan, CL; Andrews, JG; Hay, JG. Selection of body segment parameters by opti-
mization methods. Journal of Biomechanical Engineering , 1982 104(1), 38–44.
[66] Veldpaus, F; Woltring, H; Dortmans, L. A least-square algorithm for the equiform

transformation from spatial marker coordinates. Journal of Biomechanics, 1988 21,
45–54.
[67] Wicke, J; Dumas, GA; Costigan, PA. A comparison between a new model and current
models for estimating trunk segment inertial parameters. Journal of Biomechanics,
2009 42(1), 55–60.
[68] Winter, DA. Biomechanics and motor control of human movement. Wiley inter-
sciences, second edition, 1990.
[69] Woltring, H. Data processing and error analysis. 1990. In Cappozzo & Berme.
[70] Woltring, H. Representation and calculation of 3D joint movement. Human movement

Sciences, 1991 10, 603–616.
INDEX
A
abduction, 353, 355, 362, 363, 374, 376 adipose tissue, 214
abnormalities, viii, 137, 138, 140, 149, 154, 164, adrenaline, 176, 200
188, 200, 288, 289 adsorption, 256
acceleration, 21, 23, 28, 30, 52, 235, 243, 370, adults, xii, 136, 159, 198, 228, 229, 231, 243,
376 244, 256, 289, 291, 293, 301, 315, 316
accommodation, 141, 161, 168, 182, 186, 197, advanced glycation end products, 155, 161
199, 203 aetiology, 2, 11, 48
accounting, 36, 346 AFM, 307
accuracy, viii, xi, xii, 13, 29, 35, 57, 58, 71, 89, African Americans, 106, 261
90, 91, 102, 103, 253, 258, 270, 272, 281, 291, ageing, 65, 103, 151
315, 316, 321, 353, 355, 356, 362, 364, 367, agents, 154, 198
377, 378 AGEs, 151, 154
acetabulum, 284, 289 aging, viii, 108, 116, 158, 161, 163, 164, 165,
acetylcholine, 176, 200, 219 180, 216, 264
achalasia, 188, 201 agonist, 234
acid, 159, 188, 307 aid, 44, 191, 244
acidosis, 139 air, 53, 99, 100, 101, 188, 255
acoustic, x, 49, 128, 267, 270, 273, 276, 278, 279 airplanes, 50, 55
acoustic emission, 128 albumin, 128
acromegaly, 260 alcohol, 328
acromion, 374 alcoholics, 157
actin, 299 algorithm, 285, 287, 356, 360, 361, 368, 373,
action potential, ix, 207, 209, 210, 211, 212, 214, 376, 382
216, 217, 221, 223, 228, 229 alloys, 122, 123, 125, 126, 129, 133
activation, ix, x, xi, 110, 207, 215, 224, 225, 226, allylamine, 306
227, 233, 234, 241, 243, 245, 247, 267, 270, alpha, 177, 201, 217
272, 295, 297, 299, 300, 302 alternative, 32, 75, 354, 368, 377
acute, 44, 194, 221, 224, 226, 296, 299, 303 alters, 151, 153
adaptation, viii, 111, 117, 120, 121, 150, 157, aluminum oxide, 123
162, 206, 225, 226 American Psychological Association, 54
adduction, 353, 355, 362, 363, 374, 376 amino groups, 151
adductor longus, 54adhesion, xi, xii, 59, 96, 97, amplitude, 18, 112, 140, 141, 142, 212, 213, 214,
119, 120, 126, 127, 129, 135, 305, 306, 307, 215, 216, 218, 219, 220, 221, 222, 224, 225,
308, 310, 311, 312 228, 230, 238, 272, 276, 277, 297, 299, 300
adhesion force, xii, 306, 307, 308 Amsterdam, 379, 381
adhesion strength, 127, 311 AMT, 33
386 Index
amyotrophic lateral sclerosis (ALS), 223, 224 athleticism, 39, 44, 45, 48
anatomy, 7, 17, 107, 164, 176, 193, 203, 353, 354 atomic force microscope, 308
angular velocity, 245, 275, 276, 277, 278, 279, atrophy, 122, 144, 224
362, 363, 376 atropine, 165, 204
animal models, 37, 130, 149, 159, 178 attachment, xi, 29, 61, 124, 125, 128, 134, 282,
animal studies, xii, 140, 306, 307 286, 287, 291, 331, 332
animals, 61, 124, 140, 142, 144, 150, 154, 164, attitudes, 9, 10, 22, 27, 34
168, 185, 231 autonomic neuropathy, viii, 137, 138, 155, 156,
anisotropy, 59, 63, 65, 68, 74, 80, 81, 97, 98, 103, 157, 159
105, 109, 110, 177 autonomic pathways, 157
ankle joint, 241, 247, 368 avascular necrosis, 290, 294
ANOVA, 241, 316, 337, 339, 356, 362 axonal, 219
antagonist, 234, 244, 271 axons, 216, 217, 223
antagonistic, 215
B
antagonists, 243
anthropological, 49 back pain, xi, 26, 231, 295, 296, 298, 299, 301,
anthropometry, 10, 53 302, 303
antibiotics, 307 bacterial, 138, 139
antioxidant, 155, 194 barriers, 10, 18
antrum, 141, 169, 170, 198, 203 basement membrane, 151
apatite, 126, 133, 136 behavior, viii, ix, x, 15, 109, 110, 111, 112, 119,
apoptosis, 51, 149, 151, 156, 158, 201 122, 128, 131, 132, 133, 146, 152, 162, 168,
application, vii, xi, 1, 7, 10, 11, 14, 25, 36, 59, 170, 172, 185, 205, 216, 225, 228, 233, 234,
71, 73, 80, 92, 93, 101, 103, 108, 111, 117, 236, 267
119, 121, 291, 295, 318, 344, 348, 350, 353, bending, 91, 92, 93, 96, 97, 113, 114, 135, 172
366, 380 benefits, 15, 127, 258, 302
aqueous humor, 61 biaxial, 173
Areva, 125, 128, 134, 135 biaxial orientation, 173
argument, 2 biceps, 215, 230, 231, 270, 271, 272, 300
arrest, 12, 243 biceps brachii, 215, 230, 231
arson, 158 biceps femoris, 300
ART, 54 bile, 139, 175, 177
arteries, 151, 172 bile acids, 139
artery, 161, 199 bile duct, 177
arthroplasty, xi, 281, 289 bioactive materials, 125, 127
arthroscopy, 284 biocompatibility, 121, 122, 123, 306, 307
articular cartilage, 284, 293 biocompatible, 124, 311
articulation, 293 biocompatible materials, 311
artistic, vii, 1, 2, 3, 6, 9, 10, 11, 14, 15, 22, 24, biodegradable, 307
26, 38, 43, 49, 50 bioengineering, vii
aseptic, 292, 306 biofeedback, 14, 224, 227, 228, 230
assessment, 27, 28, 31, 32, 53, 59, 81, 84, 99, biofeedback training, 228
105, 115, 152, 155, 159, 176, 219, 247, 257, bioglass, 131, 132, 133, 134, 136
263, 282, 285, 286, 287, 288, 299, 307, 318, biological processes, 33, 37
325, 330, 344, 349, 381 biological responses, 127
assessment techniques, 28 biological systems, vii, 30
assimilation, 13 biomaterial, 122, 129, 307
assumptions, 58, 119, 168, 255, 351, 352, 368 biomaterials, vii, viii, 111, 112, 122, 125, 129,
astigmatism, 104 136
asymptomatic, 287, 288, 300, 302 biomedical applications, 122, 124
asymptotic, 186 biophysics, 280, 281
athletes, 2, 11 bipolar, 212, 213
athletic performance, 54 blood, 157, 185, 188, 192, 193, 196, 197, 229
Index 387
blood flow, 192, 193 cell surface, 151

blood glucose, 157, 185, 188 central nervous system, ix, 233, 234, 236
blood vessels, 196, 197 centrifugal forces, 307
body mass, 288, 289, 293, 352 ceramic, 122, 123, 124, 125, 131, 132, 133, 134,
body mass index, 288, 289, 293 135
body weight, 239, 241, 243, 288, 289, 290, 328 ceramics, 122, 124, 125, 130, 131, 134, 136
bolus, viii, 138, 163, 165, 166, 170, 172, 175, cerebral cortex, 230
178, 192, 203 cerebral palsy, vii, ix, 230, 233, 234, 235, 236,
bonding, xii, 124, 125, 126, 130, 132, 133, 134, 238, 243, 244, 245, 247, 248, 249
136, 306, 307, 311 chemical composition, 176
bone density, 31, 116 chemical stability, 125
bone graft, 127, 129 chemicals, 172
bone loss, 113, 119, 129 chest, 139, 182, 202, 238
bone mass, 116, 117, 133 chewing, 112, 131, 132, 133
bone mineral density, 135 childhood, 174, 288
bone remodeling, 117, 124 children, vii, ix, x, xii, 186, 194, 230, 233, 234,
bone resorption, 119, 126 235, 236, 237, 238, 239, 240, 241, 242, 243,
boundary conditions, 67, 90, 91, 103, 190 244, 245, 246, 247, 248, 249, 256, 265, 293,
bovine, 105, 107, 108, 172, 264, 266 315, 316, 319, 348, 380
bowel, 138, 150, 156, 157, 178, 179, 185, 188, chloride, 209
195, 196, 202 cholinergic, 141, 157, 177, 198
brain, 53, 140, 209, 217, 218, 219, 220, 226, 227, chromium, 123, 133
230, 235 chronic disease, 138
brain functions, 220 chyme, 141, 158, 165, 172
circulation, 151, 229
C
classical, 3, 4, 29, 39, 45, 47, 48, 367, 368, 371,
cadaver, 176, 192, 380 372, 374
caecum, 139, 175 classical mechanics, 29
calcium, 125, 128, 136, 306 classification, 116, 122, 226
calibration, 355 clinical neurophysiology, 230
Canada, 1, 50, 52, 231, 326 clinical symptoms, 288
capillary, 253 clinical trials, x, xi, 109, 281, 282, 291
capsule, 107, 124 closed reduction, 288
cardiac arrest, 12 clusters, 140
cartilage, 116, 282, 284, 286, 288, 289, 292, 293 CO2, 133, 310
cartilaginous, 284 coatings, xi, 124, 125, 126, 127, 128, 129, 130,
caspase, 151 131, 132, 133, 134, 135, 305, 306
CAT, xi, 282, 284 cobalt, 123, 133
CAT scan, xi, 282, 284 coefficient of variation, 211
catalyst, 50 cognition, 12
cataplexy, 228 coherence, 253, 260, 347
categorization, 43 cohesion, 49
catheter, 166 coil, 217, 220, 226
Caucasians, 106, 261, 353 colitis, 161, 179, 186, 200, 202, 204
Celiac disease, 139 collaboration, 11, 50
cell, vii, ix, xi, xii, 119, 124, 150, 151, 159, 161, collagen, vii, 52, 59, 60, 61, 63, 70, 80, 97, 102,
207, 209, 211, 219, 305, 306, 307, 308, 310, 105, 106, 107, 109, 116, 144, 149, 151, 153,
311, 312 159, 161, 168, 172, 179, 180, 181, 185, 186,
cell adhesion, xi, xii, 305, 306, 307, 308, 311, 195, 196, 197, 203, 252, 254, 259
312 colon, 139, 142, 149, 150, 156, 157, 158, 161,
cell culture, 124, 308 174, 177, 186, 187, 188, 190, 191, 194, 195,
cell death, 151 196, 199, 202, 204
cell line, 310 colostomy, 186, 199
388 Index
combined effect, 69 coping strategies, ix, 234, 237, 244, 246, 247
communication, 10, 38, 43 copper, 268, 307
community, 12, 158 correlation, xi, 63, 89, 148, 156, 185, 236, 256,
compatibility, viii, 111, 121, 132, 348 265, 281, 286, 334, 335, 336, 339, 344, 346,
compensation, 3, 23, 246 347
competency, 12 correlation coefficient, 148, 334, 335, 336, 344
competition, 53 correlations, 89, 106, 259, 318, 324, 335, 338
complement, 164 corrosion, 122, 123, 125
complex systems, 29 cortex, ix, 207, 209, 217, 218, 219, 221, 226,
complexity, x, 15, 38, 91, 102, 164, 190, 192, 227, 228, 229, 230, 231
193, 276, 279, 281 cortical stimulation, 214, 228
compliance, 145, 159, 168, 174, 175, 182, 183, corticospinal, 208, 214, 218, 219, 221, 227, 229,
186, 188, 189, 198, 202 230
complications, 110, 138, 158, 160, 306, 321 counterbalance, 235, 237, 241
components, viii, 18, 36, 58, 63, 65, 72, 103, 111, coupling, 207, 218, 219
118, 119, 142, 151, 153, 165, 167, 176, 282, covalent bond, 123
283, 362, 373, 376 CRC, 53, 130, 132, 136
composites, 128, 129 creep, 59, 63, 74, 84, 85, 86, 107, 151, 159
composition, 63, 103, 139, 176 critical points, 24
comprehension, vii, 1 cross-cultural, 49
computation, 172, 173, 283, 285, 286, 287, 288, cross-linking, 63, 65, 109, 151, 159
368 cross-sectional, 67, 135, 143, 144, 152, 165, 176,
computed tomography, 37, 190, 283, 287 180, 182, 183, 185, 188, 189, 190, 258, 288,
concentration, 73, 92, 114, 172, 185 293, 299
conceptualization, 50 crosstalk, 211, 225
conditioning, 218, 219, 296 CT scan, 292, 353
conduction, 211, 216, 221, 223, 224, 226, 227, cultural character, 11
229, 231 cycles, 32, 33, 37, 127, 177, 356, 374
conductivity, 18, 123 cycling, 25, 215, 381
conductor, 211 cytoskeleton, 119, 120
confidence, 334
D
configuration, 151, 153, 173, 178, 181, 200, 203,
212, 213, 282, 356 daily living, 215, 300
confocal laser scanning microscope, 309 data analysis, 17, 330
congenital hip dislocation, 288, 292 data collection, 21, 332, 333, 380
connective tissue, 115, 124, 125, 151, 153, 159, decisions, 244, 245
259 decomposition, 379
consensus, 251 defecation, 139, 177, 203
consent, 325, 328 deficiency, 149, 182
conservation, 245 deficits, ix, 224, 233, 238, 244
constipation, viii, 137, 139, 142, 186, 194 definition, 14, 98, 99, 121, 190, 201, 237, 322,
constraints, ix, 233, 236, 244, 277, 280, 369, 377, 345, 357, 359, 371, 376, 378, 379
381 deformation, 60, 61, 63, 65, 66, 71, 77, 82, 84,
construction, 59, 74, 89, 90, 91, 98, 103, 321 90, 91, 93, 96, 97, 101, 119, 122, 144, 145,
contractions, 18, 52, 140, 141, 142, 166, 170, 152, 166, 170, 172, 177, 189, 196, 254, 256,
176, 177, 198, 203, 208, 212, 215, 221, 223, 284, 356, 379
226, 227, 228, 229, 230, 231, 236, 299, 300 deformities, 320, 321
control condition, 318, 324, 344 degradation, 52, 71
control group, 143, 150, 183, 186, 289, 320 degrees of freedom, 277, 356, 368, 377
convection, 310 delayed gastric emptying, 141, 145, 182
convex, 255 denervation, 223
COP, 235, 236, 237, 239, 240, 241, 243, 244, Denmark, 137, 163
246, 247, 248
Index 389
density, 31, 116, 135, 149, 159, 214, 221, 222, drugs, ix, 12, 163, 172, 231
252, 256, 308, 375, 376 drying, 71
dental implants, 113, 124, 125, 127, 128, 129, duodenum, 141, 145, 146, 147, 149, 159, 170,
130, 132, 133 171, 173, 174, 175, 176, 177, 183, 185, 196,
dentistry, 112, 125 197, 199, 202, 203
Department of Health and Human Services, 131, duration, 18, 30, 31, 32, 33, 34, 53, 71, 112, 140,
301 141, 142, 146, 149, 156, 182, 185, 214, 215,
dependent variable, 246, 334, 339 216, 219, 224, 239, 241, 243, 300, 373
depolarization, 209, 216, 221 dynamic control, 49
deprivation, 110 dynamic environment, 42
descending colon, 174 dyspepsia, ix, 163, 183, 194, 203
detachment, 119, 254, 262, 311 dysphagia, 139, 188, 204
detection, 210, 211, 213, 216, 225 dysplasia, xi, 281, 287, 288, 289, 291, 293, 294
deviation, 62, 66, 95, 214, 235 dystonia, 35
diabetes mellitus, viii, 137, 142, 154, 155, 156, dystrophin, 51
157, 158, 159, 160, 161, 162, 178, 196, 202,
E
205, 321
diabetic ketoacidosis, 139 ears, 49, 288
diabetic neuropathy, 158 ECM, 307
diabetic patients, 138, 139, 140, 141, 142, 145, Education, 51, 55, 315, 317
152, 154, 157, 160, 161, 180, 182, 188, 203 EEK, 308
diarrhea, viii, 137, 138, 139, 141, 155 elastic deformation, 122
dichotomy, 2 elasticity, 105, 108, 109, 122, 151, 166, 175, 186,
diet, 150, 174, 179, 184 204, 252, 254, 255, 258, 263, 264
differentiation, 120, 129, 200 elastin, vii, 168, 196
digestion, 164, 165, 196 elbow, x, 23, 24, 25, 26, 35, 228, 231, 267, 269,
digestive tract, 165 270, 271, 272, 275, 276, 277, 278, 364, 366,
diodes, 16 375
disability, 12, 226, 303 elderly, 319, 320, 321, 348, 349
discomfort, 139, 321, 325, 349 electrode placement, 221, 225
discourse, vii, 1, 2, 3, 12, 14, 26, 44 electrodes, 18, 208, 209, 211, 212, 213, 216, 224,
discs, 296, 307, 308 225, 270
disease model, 151, 154 electrolytes, 139, 165
disease progression, 102 electromagnetic, 49, 217, 218
diseases, ix, 103, 153, 163, 164, 179, 189, 223, electromyogram, ix, 207, 208, 211, 228, 270, 299
224, 252, 253 electromyography, 15, 208, 209, 225, 226, 227,
dislocation, 288, 292, 293, 356 228, 229, 234, 302
dislocations, 306 electron, 154
disorder, 154, 182, 188, 228, 238 electrophysiological properties, 221
dispersion, 166, 286, 287 electrophysiology, 159
displacement, 35, 255, 258, 269, 293 elongation, 67, 69, 88, 166
disposition, 158 empathy, 38, 42
distraction, 106 encapsulation, 119, 126
diversity, 29, 36 encouragement, 270
DNA, 150 endocrine, 149, 160
donor, 65, 77, 79, 81, 89 endothelial cell, 161
dopamine, 219 endothelium, 59, 61, 63, 89, 90, 95, 96, 107, 251
dopaminergic, 219 energy, 21, 44, 45, 245, 246, 380, 381
dorsi, 244 environment, viii, 9, 21, 50, 51, 112, 119, 120,
drug delivery, 124 124, 135, 137, 164
drug delivery systems, 124 environmental factors, 174
drug therapy, 293 enzymatic, 151, 154, 160, 204
drug-induced, 159 epidemiology, 34, 154
390 Index
epidermal growth factor, 200, 205 fecal impaction, 139

epiphysis, xi, 281, 290, 294 feces, 165, 188
epithelium, 59, 61, 63, 74, 81, 82, 83, 89, 90, 95, feedback, 117, 224, 273
96, 106, 125, 151, 251 feeding, 184, 195
equilibrium, 89, 113, 286 feelings, 2
ergonomics, 208 feet, xii, 4, 44, 235, 239, 317, 318, 320, 321, 322,
esophagitis, 188, 194 323, 328, 338, 345
esophagus, 138, 139, 140, 141, 143, 144, 155, females, 320
157, 159, 161, 165, 166, 167, 175, 179, 180, femoral neck, xi, 281, 290
181, 182, 188, 194, 196, 197, 198, 199, 201, femur, 127, 239, 283, 297
202, 204, 205 fetal, 310
estimating, 35, 52, 353, 355, 368, 378, 382 FFT, 18
etiology, 27, 34, 154, 182, 290, 292 fiber, 116, 122, 124, 128, 129, 132, 135, 151,
euglycemia, 141 208, 210, 211, 214, 216, 221, 222
Europe, 109 fibers, ix, 61, 116, 151, 168, 194, 207, 209, 211,
evoked potential, 216, 217, 218, 220, 226, 230 212, 216
evolution, 377 fibrillar, 60, 107
examinations, 37, 223 fibrils, 59, 60, 63, 67, 70, 80, 97, 98, 102, 105,
excitability, 142, 216, 217, 219, 221, 228 106, 108, 252, 259
excitation, 207 fibrinogen, 128
excitatory postsynaptic potentials, 216, 217 fibula, 347
exclusion, 83 filament, 119
execution, 21, 91 film, 39, 42, 61, 81, 253
exercise, xi, 32, 54, 91, 93, 135, 208, 214, 226, filters, 211, 364
227, 231, 295, 296, 297, 298, 299, 300, 301, financial resources, 285
302, 303 finite element method, 109, 110, 136
exertion, 23 Finland, 111
expertise, 15, 53 fires, 272
exposure, 39, 40, 41, 42, 103, 136, 188, 287, 288, fish, 208
289 fitness, 298
extensor, 229, 244, 283, 296, 297, 298, 299, 300, fixation, xi, 125, 127, 135, 295, 296, 297, 299,
301, 302 300, 306
external fixation, 297 flare, 322
external validation, 9 flexibility, 122, 188
extracellular matrix, 59, 60, 104, 110, 151, 158, flexor, 228, 229, 231, 243, 244
159 flexural strength, 125
extraction, 73, 124, 228 flow, 71, 153, 164, 172, 175, 189, 192, 193, 194,
extraocular muscles, 60, 107 195, 201, 307, 308
extrusion, 284 flow rate, 175, 307
eye, vii, 57, 58, 61, 66, 67, 81, 89, 94, 103, 104, fluctuations, 230, 231
106, 138, 218, 251, 253, 255, 257, 258, 260, fluid, 61, 177, 307, 308, 310, 311
266 fluid transport, 61
fluoroscopy, 382
F
focusing, vii, 1, 57, 348
factor analysis, 323 folding, 200
factorial, viii, 137, 138, 154 food, viii, 112, 132, 138, 139, 141, 163, 165, 172,
failure, 9, 26, 38, 73, 119, 122, 124, 125, 126, 178, 183, 193, 205
140, 222, 225, 306, 311 Food and Drug Administration (FDA), 126, 127,
family, 268 131
fasting, 141, 179, 183, 195, 203 footwear, xii, 20, 317, 318, 319, 320, 321, 322,
fat, 18, 151 323, 324, 325, 330, 343, 344, 345, 346, 347,
fatigue, 118, 119, 122, 127, 221, 222, 226, 227, 348, 349, 350
228, 229, 230, 297, 300 Ford, 174, 196
Index 391
fracture, 114, 117, 119, 122, 124, 125, 127, 134 gestures, 11, 13, 43, 44, 47, 48
fractures, 114, 117, 119, 124, 306 Ghrelin, 159
fragility, 116 Gibbs, 112, 131
France, 4, 201, 208, 231, 350, 351, 374 girth, 327, 333, 334, 338, 342, 344, 345, 346, 347
FRC, 122, 124 glass, 124, 125, 126, 127, 128, 129, 130, 131,
freedom, 3, 277, 356, 368, 374, 377 132, 133, 134, 135, 308, 309, 310
friction, 29, 321 glasses, 125
FSA, 326, 332, 333 glaucoma, 103, 107, 110, 251, 254, 256, 261,
functional changes, 154, 183 262, 265
fundus, 141, 157, 168, 169, 170, 198 glial, 159
glucagon, 150, 155
G
glucose, 157, 185, 188
GABA, 219, 229, 231 gluteus maximus, 296
GABAB, 220 glycation, 151, 154, 155, 160, 161
GABAergic, 219 glycoproteins, 59
gait, xi, 39, 224, 230, 231, 244, 245, 246, 247, glycosaminoglycans, 252
248, 249, 282, 283, 287, 291, 292, 325, 352, goals, 1, 5, 10, 14, 49, 352
378, 380, 381, 382 gold, 114, 253, 375
gallbladder, 155, 175 gold standard, 253
ganglia, 159 gravitational force, 269
gas, 177, 188 gravity, x, 20, 48, 220, 235, 240, 267, 269, 271,
gastric, 139, 140, 141, 144, 151, 155, 156, 157, 370
158, 159, 161, 168, 169, 170, 172, 175, 176, growth, viii, xi, 58, 116, 126, 150, 155, 160, 163,
182, 183, 188, 190, 192, 193, 194, 195, 197, 164, 165, 177, 178, 179, 180, 183, 186, 193,
198, 199, 200, 201, 202, 203 195, 198, 199, 200, 204, 205, 225, 294, 295,
gastric mucosa, 144, 161 299
gastric outlet obstruction, 139 growth factor, 200, 204, 205
gastric relaxation, 201 growth hormone, 186, 195
gastrin, 195 guidance, 247, 270, 280
gastrocnemius, 243 guidelines, 127, 322, 345, 348
gastro-esophageal junction, 201 gut, 155, 157, 158, 165, 183, 186, 188
gastroesophageal reflux disease, 140, 182 H
gastrointestinal, vii, viii, 137, 139, 154, 155, 156,
157, 158, 159, 160, 161, 162, 163, 164, 189, hallux valgus, 320
195, 196, 202, 203 hamstring, 235, 296, 298, 300, 302
gastrointestinal tract, vii, 137, 154, 156, 157, 158, hands, 16, 270
159, 161, 196 head injuries, 53
gastroparesis, 139, 144, 155, 156, 157, 159, 160, healing, 103, 117, 119, 127, 128, 254, 266
182, 194, 199, 200 health, xi, 2, 9, 10, 12, 15, 51, 55, 138, 160, 164,
gauge, 216 189, 191, 194, 200, 201, 205, 226, 252, 293,
gel, 126, 128, 131, 134, 135 295, 324, 348, 349
gender, 54, 202, 256, 289, 290, 320 Health and Human Services, 131, 301
gene, 27 health problems, 55, 138
general education, 6 health status, 293
generalization, 29, 346 healthcare, 138, 222
generalizations, 27 heart, 138, 139
generation, viii, 20, 49, 109, 137, 208, 214, 237, heart disease, 138, 139
243, 246, 263, 307 heat, 159, 182
generators, 282 height, 24, 25, 131, 151, 235, 239, 282, 283, 287,
genetic factors, 259 319, 322, 323, 325, 328, 330, 334, 352
geography, x, 267 hemiplegic, 230, 236
geometrical parameters, 287, 290 hernia, 188
geriatric, 298, 302 herniated, 301
392 Index
heterogeneous, 96, 247 hypertension, 254, 261, 262, 265

hiatal hernia, 189 hypertensive, 254, 261, 262
high resolution, 258 hypertrophy, 144, 151, 157, 301
high-frequency, 199, 201, 205, 219 hypothesis, 99, 101, 120, 180, 288, 289, 290,
high-level, 13 375, 377
high-speed, 39 hysteresis, x, 59, 74, 79, 89, 98, 99, 101, 102,
hip arthroplasty, 289 103, 105, 107, 174, 251, 252, 256, 264, 265,
hip endoprostheses, 306 266
hip joint, 130, 235, 237, 241, 242, 243, 244, 282,
I
283, 284, 286, 289, 290, 291, 292, 293, 353,
355, 357, 358, 362, 364, 366, 367, 378, 379, iatrogenic, 258
381 ICC, 142, 156, 335, 336, 344
hip replacement, 132, 293 identity, 360, 371
hips, xi, 281, 284, 286, 287, 288, 289, 290, 291, idiopathic, xi, 188, 281, 288, 289, 291, 293
292, 293, 294, 296 IGF, 157
Hispanics, 106, 261 IGF-I, 157
histamine, 176, 200 ileum, 145, 146, 170, 172, 173, 174, 175, 183,
histochemical, 154, 158 185, 199, 203, 204
histological, 37, 124, 131, 134, 149, 180, 183, IMA, 152
187 imagery, 40, 42, 49, 227
histology, 128, 136 images, 15, 37, 38, 39, 44, 60, 190, 192, 258, 299
hockey, 349 imaging, xi, 37, 108, 152, 164, 168, 174, 176,
holistic, 21, 36 190, 193, 196, 197, 201, 255, 258, 259, 264,
Holland, 182, 198, 379 282, 284, 285, 287, 291, 302, 353, 354, 378
hologram, 258 imaging modalities, 164
homeostasis, viii, 136, 137, 151, 152, 189 imaging techniques, 190, 193, 287, 291
homogenous, 58, 83, 96, 115, 172 immobilization, 129, 226
Hong Kong, 317, 323, 349 immunohistochemical, 154, 161
hormone, 135, 151, 186, 195 impairments, ix, x, 233, 234, 237, 244, 246, 247
horses, 60 implantology, 124, 125, 130
host, viii, 111, 121, 125 implants, vii, viii, xi, xii, 111, 112, 113, 114, 118,
host tissue, viii, 111, 121 119, 122, 123, 124, 125, 126, 127, 128, 129,
human experience, 38, 42, 49 130, 131, 132, 133, 134, 135, 136, 305, 306,
human gait, 231, 380 307
Human Kinetics, 51, 227, 229, 350, 382 impulsive, 275, 276
human subjects, 16, 230 in situ, 156, 260, 262, 263, 265, 266
humans, ix, 17, 18, 61, 142, 144, 155, 160, 164, in vitro, 124, 128, 132, 143, 145, 153, 165, 168,
168, 203, 204, 207, 217, 227, 229, 230, 231, 174, 177, 188, 198, 203, 204, 205, 307
252 in vivo, 66, 89, 109, 124, 127, 132, 133, 134,
hyaline, 116 135, 151, 153, 165, 168, 170, 174, 186, 190,
hydration, 65, 71, 73, 107, 256, 260 191, 193, 199, 200, 204, 216, 217, 224, 227,
hydrodynamic, viii, 138, 163, 178, 307 254, 255, 258, 264, 283, 367, 368, 381
hydrogen, 141, 221 inactive, 125, 215
Hydrothermal, 129 incidence, 26, 258, 288, 289, 292, 320, 321
hydroxyapatite, 126, 127, 130, 131, 132, 133 inclusion, 103
hydroxylapatite, 125, 133 incompressible, 66, 84, 105
hyperalgesia, 156 incubation, 310, 312
hyperglycemia, 138, 142, 152, 155, 157, 159, 202 incubation period, 310
hyperinsulinemia, 142 independent variable, 318, 334, 339
hyperopia, 262 indication, 62, 272, 289
hyperreactivity, 140 indicators, 23, 48, 333
hypersensitive, 182, 202 individual characteristics, 321
hypersensitivity, 186, 188, 199 individual differences, 324
Index 393
induction, 143, 149, 151, 188, 216, 217, 218 internal processes, 14
industrial, 2, 20, 283 interneuron, 207
industry, 26, 50, 131, 319, 322 interstitial, 144, 156, 157, 158, 198
inertia, 2, 10, 48, 352, 353, 354, 368, 369, 371, interval, 175, 214, 216, 218, 219, 246
372, 374, 375, 376, 377, 378, 380, 381 intervention, 12, 220, 245, 252, 300, 321
infancy, 9, 26 intestinal obstruction, 185
infants, 248 intestine, 139, 140, 141, 145, 146, 151, 154, 155,
infection, xi, 125, 305, 306 157, 158, 159, 160, 162, 165, 170, 171, 172,
infections, xi, 305, 306, 307 174, 183, 184, 185, 186, 190, 192, 194, 195,
infectious, xii, 306, 311, 312 196, 197, 199, 200, 203, 204, 205, 206
inflammation, 178, 187 intonation, x, 19, 268, 270, 273
inflation, 65, 66, 67, 68, 71, 72, 73, 74, 76, 85, intracortical microstimulation, 221
101, 109, 166, 175 intramuscular, viii, 137, 142, 160, 209, 211, 212,
informed consent, 325 216, 222
infrared, 239, 374 intraocular, x, 66, 71, 74, 76, 91, 99, 100, 101,
ingestion, 141 107, 108, 109, 110, 251, 252, 255, 259, 260,
inhibition, 177, 216, 217, 218, 219, 229, 230, 231 261, 262, 263, 264, 265
inhibitor, 158 intraocular pressure, x, 66, 71, 74, 76, 91, 99,
inhibitory, 142, 156, 177, 198, 218, 219 100, 101, 107, 108, 109, 110, 251, 252, 259,
inhibitory effect, 142 261, 262, 263, 264, 265
inhomogeneities, 211 intrinsic, 152, 189, 268, 300
injection, 165 invasive, x, 18, 203, 217, 220, 226, 253, 281,
injections, 161 284, 367, 377
injuries, 2, 3, 10, 11, 22, 25, 26, 27, 30, 33, 34, inversion, 181
35, 37, 52, 53, 55, 89 Investigations, 130, 168
injury, vii, 1, 2, 3, 9, 10, 11, 12, 13, 15, 20, 22, investigative, 216
26, 27, 30, 31, 32, 33, 34, 35, 36, 51, 52, 55, ionic, 123, 164, 217
57, 58, 61, 89, 142, 172, 178, 220, 224, 225, ions, 211, 306, 312
231, 325 IOP, 74, 75, 76, 90, 91, 100, 102, 103, 107, 252,
innervation, 149, 156, 198, 211, 213 253, 254, 255, 25iron, 122
innovation, 38, 106, 132, 137, 163 irradiation, 186, 198
insertion, 61, 216, 282, 283 irritable bowel syndrome, 188, 196, 202
insight, 25, 34, 216, 222, 223, 236, 348 IS, 108, 263
instability, 228, 246 ischemia, 221, 227, 229, 231
instruction, 4, 5, 6, 10, 14 ischemic, 52, 139, 230
instruments, 5, 20, 25, 49, 53, 253, 258, 259, 268, ischemic heart disease, 139
272 isolation, 297, 301
insulin, 138, 150, 154, 155, 156, 157, 158, 193, isotropic, 97, 98, 116
195, 198, 262 isotropy, 80, 84
insulin dependent diabetes, 158 iteration, 373
integration, xi, 1, 22, 36, 38, 106, 305, 306
J
integrin, 312
integrins, 312 jaw, 35, 112, 118, 129
integrity, 103, 127, 133, 186, 217, 295 jejunum, 145, 146, 173, 174, 183, 185, 199, 203
interaction, viii, 5, 13, 20, 53, 97, 119, 120, 121, joints, 16, 21, 22, 23, 34, 35, 118, 235, 236, 239,
138, 153, 163, 178, 235, 307, 318 241, 242, 243, 244, 246, 247, 282, 296, 353,
interactions, 17, 21, 22, 23, 164, 356 354, 355, 356, 357, 364, 368, 380
interface, viii, 5, 10, 17, 49, 50, 53, 83, 111, 112,
118, 119, 122, 124, 125, 126, 127, 129, 130, K
131, 132, 133, 134, 135, 151, 182, 307, 318,
321, 325, 326, 332, 343, 347 keratinocytes, 59
interfacial bonding, 130 keratoconjunctivitis, 260
interference, 213, 214, 215, 221, 222, 310, 333 keratoplasty, 109, 262
ketoacidosis, 139
394 Index
kidney, 138, 156 201, 213, 214, 254, 277, 282, 286, 287, 307,
kinematics, xii, 16, 52, 53, 54, 236, 242, 244, 308, 322, 376, 381
245, 246, 282, 350, 351, 352, 353, 354, 355, linear function, 175
356, 357, 362, 364, 366, 367, 368, 375, 376, linear regression, 185, 322
377, 380, 381, 382 links, 21, 38, 49, 83, 151, 160, 378
kinetic parameters, 235 locomotion, vii, ix, 1, 233, 236
kinetics, 231, 239, 244, 245, 246, 247, 248, 350, London, 53, 104, 107, 129, 130, 134, 135, 156,
351, 352, 356, 375, 380 198, 230, 231, 248
knee, xi, 124, 213, 229, 234, 235, 237, 239, 241, long period, 127
242, 243, 244, 245, 247, 248, 283, 293, 305, longevity, 2, 15, 112, 118, 127
306, 353, 364, 368 longitudinal studies, 225
knee replacement, 124 low back pain, xi, 231, 295, 296, 298, 299, 301,
302, 303
L
lower esophageal sphincter, 140, 175, 194, 195,
lactose intolerance, 157 201, 205
lamella, 60, 63, 89 low-level, 229
lamellae, 60, 61, 63, 82, 83, 84, 96, 102, 106, 252 LSM, 309, 310
lamellar, 59, 83, 96, 109, 116 lumbar, xi, 227, 295, 296, 297, 298, 299, 300,
lamina, viii, 97, 110, 137, 159, 160, 161, 307, 301, 302, 303
310 lumbar spine, 296, 303
laminar, viii, 97, 137, 159, 160, 161, 307, 308, lumen, 153, 164, 165, 170, 176, 180, 201
310 luminal, 139, 144, 151, 164, 165, 170, 172, 174,
laparoscopic gastric, 197 176, 183, 185, 190, 200
large intestine, 165, 174, 186, 197, 204 lung, 189
laser, 71, 72, 109, 126, 133, 254, 257, 259, 260, M
262, 263, 265, 266, 283, 307, 309, 310, 325
laser ablation, 254, 257 magnetic, ix, 37, 164, 174, 190, 196, 207, 209,
latency, 215, 223, 224, 237, 241 216, 217, 218, 224, 226, 227, 229, 230, 231,
lateral epicondyle, 239 299, 302
lateral sclerosis, 224 magnetic resonance imaging, 37, 164, 174, 190,
law, 127, 170, 174, 177, 186, 286, 370 196, 302
learning, vii, xii, 1, 2, 3, 4, 7, 9, 10, 11, 12, 13, Maillard reaction, 158
14, 23, 24, 25, 35, 36, 48, 51, 52, 53, 55, 289, maintain balance, 239, 282
315, 316 maintenance, 116, 183, 255, 259
learning outcomes, 11 malnutrition, 178, 184
learning process, 9, 10, 11, 12, 13, 14, 23, 24, 25, management, ix, 20, 26, 34, 35, 36, 228, 233,
35, 316 237, 245, 251, 253, 254, 259, 296, 300, 378
leg, 44, 51, 52, 245, 246, 248, 282, 299, 350 mandible, 112, 116, 119, 127
lending, 44 mandibular, 115, 130
lens, 103, 107, 252, 259 mapping, 50, 59, 89, 103, 107, 168, 220, 360
lenses, viii, 57, 63, 89, 102, 259, 262, 263 market, 123, 321
lesions, 35 mastication, 112, 118, 134, 188
lifetime, 291 masticatory, 112, 128
ligament, 118, 124 matrix, 59, 60, 89, 104, 110, 117, 129, 151, 158,
likelihood, 127 159, 221, 252, 307, 353, 357, 358, 359, 360,
limbus, 61, 62, 63, 65, 66, 92, 102, 105, 106, 252 361, 362, 363, 364, 366, 367, 368, 369, 370,
limitation, 317, 323, 354, 376 371, 372, 373, 374, 375, 376, 377, 381
limitations, 15, 20, 68, 176, 223, 224, 225, 236, matrix metalloproteinase, 110
258, 289, 291, 318, 349 matrix protein, 151, 159
Lincoln, 149, 158 maturation, 6, 129
linear, 21, 44, 58, 70, 76, 77, 91, 95, 108, 122, maxilla, 116, 119
148, 165, 166, 172, 174, 175, 177, 185, 190, mechanical behavior, viii, 111, 112, 122, 131,
168
Index 395
mechanical properties, viii, 60, 63, 105, 108, 111, morning, 252
112, 115, 116, 122, 124, 125, 127, 134, 138, morphological, viii, xi, 37, 125, 130, 137, 138,
163, 164, 165, 168, 174, 175, 177, 178, 182, 142, 143, 144, 145, 149, 151, 152, 154, 165,
183, 186, 188, 191, 193, 194, 202, 204, 263 167, 174, 183, 185, 189, 190, 192, 197, 235,
mechanical stress, 118, 129, 183, 189 281, 282, 285, 291, 299
mechanical testing, 128, 129 morphology, 121, 158, 165, 175, 183, 186, 196,
median, 66, 69, 93, 214, 221, 222, 226, 287 257, 287, 353
medications, 252 morphometric, 144, 149, 162, 170, 179, 182, 183,
medicine, 12, 51, 54, 222 185, 187, 190, 198, 200, 205
men, 62, 188, 200, 228, 289, 374 mortality, 138
mesenchymal stem cell, 312 motor activity, 140, 156, 159
mesenchymal stem cells, 312 motor behavior, ix, x, 233, 234, 267
meta-analysis, 252, 259, 301 motor control, 2, 7, 10, 11, 12, 13, 14, 15, 17, 22,
metabolic, 139, 289 23, 24, 27, 34, 36, 216, 230, 231, 235, 249,
metabolic acidosis, 139 301, 303, 352, 382
metabolism, 52 motor function, viii, 137, 138, 154, 163, 189,
metabolites, 221 198, 202, 231, 234
metal ions, 312 motor neurons, ix, 207, 236
metalloproteinase, 110 motor skills, ix, 3, 11, 15, 21, 233, 234, 246
metals, 122, 124 motor system, 234
metaphors, 42, 49 motor task, 234
metatarsal, 319, 320, 322, 328, 342, 347 motorneurons, 217
metric, 375 MRI, xi, 37, 164, 174, 190, 282, 284, 300, 302
mice, 135, 141, 160, 181, 187, 204 mucosa, 142, 143, 144, 146, 151, 158, 161, 166,
microgravity, 136 167, 168, 169, 171, 174, 177, 180, 182, 183,
micrometer, 260 184, 188, 201, 205
microorganisms, 125 multidimensional, 15, 34
microscope, 106, 172, 258, 307, 308, 309 multidisciplinary, 3, 11, 15, 36
microscopy, 196, 260, 309 multimedia, 49, 352
microstructure, 59, 61, 63, 65, 80, 89, 102, 103, multiple regression, 339
130, 257 multiple sclerosis, 224
microvascular, 160 muscle atrophy, 144, 224
middle-aged, 160, 186, 198 muscle cells, 177, 198, 209
migration, 120 muscle contraction, 18, 165, 170, 176, 195, 208,
mineralization, 119 218, 219, 236, 239, 244
misleading, 27, 286 muscle force, ix, 29, 231, 233, 236, 244, 245,
mixing, 170, 195 246, 247, 283, 300
mobility, 245, 246 muscle mass, 296, 297, 299
modalities, 37, 164, 302 muscle performance, 228, 297
modality, 37 muscle relaxation, 186
modeling, 14, 15, 17, 21, 22, 23, 27, 28, 29, 36, muscle spindle, 216
55, 104, 110, 163, 193, 194, 323, 353 muscle strength, xi, 295, 300
modulation, 151, 199 muscle tissue, 37, 209, 216
modulus, viii, 64, 65, 76, 77, 84, 95, 96, 105, muscle weakness, 234, 243, 244, 245
108, 109, 111, 122, 125, 128, 151, 166, 173, muscular contraction, 221
175, 183, 185, 202, 204, 254, 255, 264, 266 muscular system, 208
molecular structure, 124 musculoskeletal, vii, ix, x, xi, 1, 13, 52, 53, 55,
molecules, 151 224, 233, 234, 236, 237, 238, 244, 246, 247,
Møller, 263 283, 295, 301, 302
momentum, 29, 30, 235 musculoskeletal system, ix, 13, 234, 237, 302
monkeys, 129, 133, 134 music, vii, 1, 2, 3, 4, 5, 6, 7, 11, 12, 24, 49, 51,
monolayer, 61 52, 53, 55, 269, 280
Monroe, 106 musicians, 2, 10, 26, 49, 51, 52, 55, 268, 279
396 Index
myofibrillar, 54 non-insulin dependent diabetes, 158

myopia, 89, 103, 110, 261, 262, 263 non-invasive, 18, 209, 217, 284
myopic, 259, 263 non-linearity, 177
myosin, 299 non-obese diabetic, 141, 160
non-uniform, 67, 74, 90, 165, 171, 180, 286, 293
N
noradrenaline, 176, 197, 200
NASA, 226 norepinephrine, 219
Nash, 64, 108, 263, 315 normal aging, 116
National Institutes of Health, 247 normalization, 18, 37, 214, 229
natural, viii, 26, 48, 58, 93, 111, 124, 254 norms, 17, 28
naturalization, 14 North America, 345
nausea, viii, 137, 139, 201 numerical analysis, 110, 201
neck, xi, 6, 35, 272, 281, 290 nutrient, 61, 81, 150, 151, 176, 199
necrosis, 290, 292, 294 O
nerve, viii, 137, 138, 145, 152, 154, 189, 209,
214, 215, 216, 217, 223, 224, 225 obese, 141, 160
nerves, 142, 177, 216, 223 obesity, 193, 293
nervous system, vii, ix, 1, 2, 188, 207, 233, 234, observations, 164, 258, 299
235, 236, 280 obstruction, 139, 178, 179, 185
network, 129, 144, 164 occlusion, 113
neuroendocrine, 155 occupational, 10, 224, 228
neuroendocrine system, 155 Ohio, 207, 225, 295
neurogenic, 188, 204 old age, 288
neurological deficit, 238 older adults, 198, 231
neurological disease, 224 older people, 350
neurologist, 222, 230 onset latency, 215, 237, 241
neuromotor, 301 open angle glaucoma, 265
neurons, ix, 149, 158, 161, 207, 217, 219, 220, operator, 37, 253, 359, 371
221, 236 ophthalmologists, 103, 251
Neuropathies, 51 optical, 16, 40, 58, 108, 172, 253, 255, 258, 259
neuropathy, viii, 107, 137, 138, 142, 154, 155, optimization, vii, 1, 24, 35, 131, 283, 356, 357,
156, 157, 158, 159, 321 358, 376, 377
neurophysiology, 229, 230 optimization method, 356, 357, 358, 377
neuroscience, 15 oral, 112, 113, 119, 124, 130, 132, 135, 160, 170
neurosurgery, 222 Oregon, 52, 239, 247, 270, 280
neurotransmission, 219 organ, 192, 199
neurotransmitter, 216, 217, 219 organic, 186
New York, 50, 51, 52, 53, 54, 55, 108, 131, 132, orthopaedic, vii, xi, xii, 126, 130, 134, 281, 282,
228, 233, 249, 267, 293 305, 306, 307, 308, 311
Newton, 105, 106, 109, 259, 353, 368, 369, 370 oscillation, 272, 277
Newtonian, 29 osteoarthritis, xi, 281, 286, 288, 289, 291, 292,
Newtonian physics, 29 293
NHS, 251 osteoblastic cells, xii, 306
Nielsen, 155, 156, 159, 186, 195, 196, 197, 198, osteoblasts, 117, 312
199, 202, 230 osteoclasts, 117
NIH, 193 osteocytes, 117, 132, 135
Nissen fundoplication, 182, 202 osteogenesis imperfecta, 181, 198
Nixon, 321, 322, 345, 350 osteopenia, 127
NMDA, 218, 219 osteosarcoma, 310
NMDA receptors, 218, 219 osteotomies, 290, 292, 294
noise, 208, 211, 215, 355, 356, 357, 362, 364, osteotomy, xi, 281, 289, 290, 291, 292, 293, 294
365, 367, 373, 377 oxidation, 151
non-enzymatic, 151, 154 oxide, 123, 125
Index 397
oxides, 125 periodontal, 118, 124

oxygen, 61, 81, 122 periodontium, 118, 135
peripheral nerve, 214, 215, 216, 217, 223, 224,
P
225
pain, viii, ix, xi, xii, 12, 20, 26, 35, 137, 139, 141, peristalsis, 161, 165, 170, 194, 197
155, 158, 159, 161, 163, 170, 174, 182, 186, permeability, 160, 203
188, 196, 202, 224, 231, 288, 295, 296, 298, perturbation, ix, 233, 234, 236, 237, 239, 240,
299, 300, 301, 302, 303, 317, 320, 321 241, 247
pain management, 300 perturbations, 243, 244, 245
pancreas transplant, 158 pH, 221, 226
pancreatic, 139, 175, 193 pharmacotherapy, 139
pancreatic insufficiency, 139 pharynx, 188
paradigm shift, 6 phenomenology, 5
parameter, 30, 32, 33, 89, 91, 190, 236, 252, 256, Philadelphia, 248, 259
264, 290, 377, 378, 380, 381 philosophical, 10
parameter estimation, 378 phosphate, 125, 126, 128, 132, 133, 306
parathyroid, 135 photon, 259
parathyroid hormone, 135 physical activity, 49, 226, 291
Paris, 5, 50, 350, 382 physical environment, 164
Parkinson, 188, 199 physical health, 9
particles, xi, 126, 131, 132, 255, 305 physical properties, 115, 123, 276
passenger, 50 physical therapy, 1, 222, 248
passive, viii, 16, 138, 144, 153, 163, 165, 168, physics, 17, 28, 29, 280
172, 175, 176, 178, 183, 201, 202 physiological, ix, 7, 10, 11, 18, 24, 26, 27, 30, 32,
patella, 353 37, 61, 112, 119, 127, 151, 153, 163, 164, 168,
pathogenesis, viii, ix, 137, 138, 142, 152, 154, 174, 175, 177, 178, 183, 198, 200, 208, 209,
159, 163, 182, 263, 301 211, 219, 221, 223, 227, 284
pathogenic, 125 physiological factors, 221
pathology, 27, 155, 222, 286, 294 physiologists, 221
pathophysiological, 178 physiology, 2, 7, 153, 158, 164, 193, 198, 203,
pathophysiology, 152, 156, 164, 193, 196, 234, 209, 217, 223, 226
244, 245 physiotherapy, 303
pathways, 152, 157, 164, 177, 182, 186, 189, piezoelectric, 105
207, 217, 224 pilot study, 134, 135
patterning, 176 pitch, x, 19, 20, 35, 267, 268, 269, 270, 272, 273,
PCT, 63, 66, 100, 102 275
pedagogical, 3, 9, 10, 11, 12, 13, 14, 15, 20, 22, placebo, 161
23, 34, 53 planar, 45, 275
pedagogy, 3, 6, 9, 10, 11, 12, 14, 20, 22, 24, 25, planning, viii, 14, 20, 57, 89, 102, 252, 290
26, 34, 35, 279 plantar, 243, 244, 322, 324, 325, 349
pediatric, 289 plasma, 126, 127, 131, 132, 134, 150, 306
PEEK, 124, 132, 308 plasticity, 220, 226, 230
pelvic, xi, 224, 227, 281, 282, 283, 287, 288, 289, platforms, 20
291, 292, 295, 296, 297, 298, 299, 300, 302 plexus, 142, 156, 158
pelvis, 186, 239, 293, 297, 300, 355, 357, 366 PMMA, 132
pendulum, 235 POAG, 256
Pennsylvania, 248 Poisson, 66, 84, 116
peptide, 150, 155 polyetheretherketone, 308
peptides, 155, 161 polyetheretherketone (PEEK), 308
perceptions, 2, 14, 15, 22, 26, 50 polyethylene, 124
perforation, 139 polymer, 124, 132, 306
performers, 4, 12, 13, 14, 15, 20, 22, 24, 26, 34, polymeric materials, 124
35, 36, 38, 269, 270, 274, 279 polymers, 122, 124
398 Index
polymyositis, 223 proteoglycans, 59, 104, 105, 307

polynomial, 75 protocol, 19, 20, 35, 119, 221, 258, 296, 298,
poor, ix, xii, 12, 15, 119, 125, 224, 233, 300, 317, 299, 318, 324, 330, 344, 354, 357, 369, 379
320, 347 protocols, 18, 245, 258, 300, 353
population, xii, 138, 261, 286, 287, 289, 291, prototype, 282
298, 321, 351, 352, 355, 381 proxy, 48, 166
pores, 107, 130 pseudo, 139, 367, 368, 380
porosity, 124, 132 psychiatric disorder, 229
porous, 130, 133 psychiatric disorders, 229
portal vein, 197 psychology, 7, 54
postoperative, 290 pubis, 355
postsynaptic, 216, 217 public health, 138
postural instability, 246 pulse, x, 99, 216, 217, 218, 219, 221, 226, 251,
posture, x, 23, 35, 44, 234, 235, 236, 243, 244, 255
247, 248, 267, 269, 275, 276, 277, 329 PUMA, 368
potassium, 209, 211, 221 pumping, 61
power, 2, 11, 39, 44, 45, 48, 58, 71, 174, 176, pylorus, 141, 176, 194
214, 221, 222, 226, 227, 231, 251, 263, 303, pyramidal, 219
346, 347, 349
Q
powers, 36
pragmatic, 303 quadriceps, 227
pre-clinical, 124, 283, 292 quality of life, 138, 160
predictability, 257 quantitative research, 2, 10, 11, 21, 22, 27, 36
prediction, 90, 323, 342, 346, 347, 378 quantitative technique, 26
predictive model, viii, 57 quasi-linear, 174, 186
predictors, xi, 226, 281, 294, 340, 341, 342, 343
press, 38, 265, 306, 349 R
presynaptic, 216, 217
prevention, vii, xi, 1, 3, 11, 12, 15, 20, 26, 27, 32, radial distance, 308
34, 35, 36, 53, 295, 301, 303, 349 radial keratotomy, 109
primary care, 349 radiation, 178, 186, 234
primary open-angle glaucoma, 262 radiation therapy, 186
primates, 107 radiography, 37, 284
proactive, 236 radiological, 289, 303
probe, 140, 166, 176, 186, 201, 253 radiopaque, 141, 157
production, ix, 20, 35, 139, 210, 233, 234, 237, radiotherapy, 186
239, 243, 244, 245, 246, 247, 272, 296, 300, radius, 61, 62, 66, 69, 90, 92, 176, 255, 283, 284,
302, 319 286, 287, 288, 364, 367
professions, 222, 325 RAGE, 155, 160
prognosis, 228, 231 random, 18, 252, 355, 362, 364
program, 98, 287, 301 ratings, 318, 323, 324, 335, 336, 337, 346
programming, vii, 57, 248 reaction time, xii, 315
prolapse, 189, 196 reactivity, 125, 182
proliferation, viii, xi, 120, 129, 137, 142, 151, real time, 49, 318, 347, 353
305, 307 reality, 1, 29, 34, 42, 57, 91
propagation, 221, 228 receptive relaxation, 176, 188
prostheses, xi, 129, 130, 133, 134, 136, 305 receptors, 151, 152, 154, 159, 161, 177, 189, 218,
prosthesis, 112, 113, 114, 122, 131 219, 229, 312
protection, 172 reconstruction, 16, 45, 47, 190, 366, 367
protein, 151, 154, 155, 179, 184, 195 recovery, 27, 32, 33, 181, 184, 220, 223, 224,
protein synthesis, 155, 195 226, 231, 236, 237, 240, 241, 242, 243, 244,
proteins, 51, 151, 158, 159, 186, 195, 199, 299, 245, 246, 248, 262, 298, 303
307 rectal prolapse, 189, 196
rectification, 18, 214
Index 399
rectum, 161, 174, 175, 177, 186, 188, 190, 195, rigidity, 118, 189, 251, 253, 254, 255, 256, 263,
196, 197, 198, 202 264, 346
rectus abdominis, 19, 20 rings, 90, 167, 170, 176, 177
recurrence, 140 risk, xi, 3, 13, 20, 26, 31, 32, 33, 34, 35, 36, 53,
red light, 332, 333 125, 164, 254, 256, 258, 262, 281, 288, 289,
redundancy, 366, 367 290, 291, 293, 294, 306, 321, 349
reference frame, 357, 361, 364, 365, 366, 369, risk assessment, 31, 32, 53, 349
370, 372 risk factors, 254, 288, 289
reflexes, 142, 160, 216, 248 risk management, 20, 34, 35, 36
refractive index, 58 risks, 2, 34
refractory, 199 robotics, 201, 378
regeneration, 36, 119 rotations, 275, 353, 380, 382
regional, 103, 168, 204 rotator cuff, 26
regression, 54, 183, 185, 322, 323, 334, 339, 340, Royal Society, 109
341, 342, 343, 346, 347, 348, 349, 353, 355, Rutherford, 2, 51
381
S
regression analysis, 183, 185
regression equation, 323, 341, 342, 346, 347, sacrum, 239
348, 353, 355 Salen, 130
rehabilitation, vii, ix, 1, 220, 222, 231, 233, 245, sapphire, 131
298, 301, 302, 303 SARA, 355
relaxation, 59, 74, 84, 85, 86, 87, 148, 151, 159, scaffolding, 124
172, 174, 176, 177, 183, 185, 186, 188, 195, scaling, 282
201, 204, 270, 299, 302 scalp, 220, 221
relaxation processes, 151 scapula, 355, 374
relaxation rate, 86 scattering, 105
relaxation time, 299, 302 Schmid, 203
relevance, 38, 49, 107, 264, 301 school, 51, 52, 55
reliability, 67, 91, 110, 318, 330, 334, 335, 344, scientific method, 7, 15
377, 380 sclera, 66, 89, 90, 93, 103, 106, 108, 110, 259
REM, 228 scleroderma, 189, 201
remediation, 12, 14, 27, 34, 36 sclerosis, 159, 178, 179, 185, 202, 224
remodelling, 63, 127, 161, 162, 164, 178, 179, sclerotherapy, 181, 197, 199
180, 184, 186, 187, 188, 189, 195, 204, 205 search, 253, 361
repeatability, 3, 20, 330, 344 searching, 35, 377
repetitions, xii, 300, 315 secondary schools, 51
research design, 225 secretion, 164
resection, 178, 179, 185, 195 segmentation, 190
reservoir, 71, 72, 172 seizure, 18
resilience, 120 self-awareness, 34
resin, 124, 129, 130 self-perceptions, 2, 26
resistance, x, 60, 61, 63, 91, 97, 107, 120, 122, self-report, 288
125, 176, 188, 189, 203, 226, 251, 252, 256, SEM, 135
264, 265, 268, 296, 297, 298, 299, 300, 301, semiconductor, 325
302, 306 sensation, 152, 170, 186, 202, 321, 330, 347
resolution, 71, 175, 201, 215, 258, 303, 353, 369, sensations, 156
376 sensitivity, 18, 94, 141, 182, 202, 258, 344, 345,
retention, 141 346, 378
retinal detachment, 254, 262 sensitization, 182
rheological properties, 194 sensors, 20, 53, 117, 270, 318, 323, 326, 330,
rhythm, 296, 300 331, 332, 333, 341, 342, 343, 347, 348, 349,
riboflavin, 109 351
sensory nerves, 141
400 Index
separation, 86, 223, 235, 246, 247 smoothness, 44

serum, 310 sodium, 209, 211
severity, 55, 140, 156, 186, 222, 287, 301 soleus, 215, 231
sex, 115, 204, 259, 321 sol-gel, 126, 128, 131, 135
shape, 13, 58, 61, 62, 90, 92, 93, 112, 118, 119, solid waste, 165
176, 211, 226, 251, 252, 255, 284, 286, 289, solidification, 364, 379, 381
291, 292, 307, 317, 319, 322, 323, 325, 345, solid-state, 270
347, 349, 375, 376 spastic, ix, 230, 233, 234, 235, 236, 238, 243,
sharing, 127 244, 245, 247, 248, 249
shear, xi, 59, 83, 84, 96, 116, 127, 143, 166, 180, spasticity, 243
281, 290, 307, 308, 311 spatial, ix, x, 21, 43, 48, 170, 175, 176, 208, 213,
shear deformation, 166 216, 220, 233, 243, 244, 267, 269, 274, 275,
shear strength, 127 276, 277, 351, 379, 382
sheep, 131, 194 spatial location, 208, 220
Shell, 104, 105 species, 60, 115, 168
shoulder, x, 23, 25, 26, 35, 45, 53, 224, 228, 267, specificity, 221, 302
269, 271, 275, 364, 366 spectrum, 122, 214, 221, 222, 226, 227, 230, 288
side effects, 139 speed, ix, 11, 23, 39, 44, 45, 48, 71, 103, 233,
sigmoid colon, 142, 174, 188, 190, 191, 196, 202 237, 243, 244, 248, 310, 316, 324
signal transduction, 117, 164 spheres, 366
signals, ix, 18, 21, 27, 29, 50, 117, 193, 207, 208, sphincter, 140, 175, 176, 177, 188, 189, 193, 194,
209, 211, 212, 213, 214, 218, 221, 223, 224, 195, 196, 197, 198, 199, 201, 202, 203, 204,
225, 229, 236, 239, 259 205
signal-to-noise ratio, 208 spinal cord, 207, 216, 217, 221, 223, 231
signs, 51, 289 spinal cord injury, 231
silica, 126, 131 spindle, 216
silicate, 135 spine, 239, 295, 296, 303
similarity, 74, 276 spondylolisthesis, 303
simulation, vii, ix, 57, 58, 59, 79, 83, 87, 91, 93, spondylolysis, 303
98, 99, 101, 102, 103, 104, 106, 128, 163, 164, spontaneous recovery, 220
205, 208, 356, 362, 379 SPSS, 333, 334
simulations, 83, 89, 94, 97, 102, 103, 199, 200, stability, ix, xi, 44, 49, 125, 130, 132, 233, 234,
349 241, 244, 245, 246, 251, 295, 299, 305, 306,
singular, 236 311
sintering, 130, 283 stabilization, 297, 298, 301, 302, 303
sites, 119, 161, 162, 206, 220, 223 stabilize, 272
skeletal muscle, ix, 52, 54, 55, 207, 209, 220, stages, xii, 10, 11, 13, 14, 25, 27, 36, 86, 99, 101,
226, 227, 230, 298, 299 286, 307, 311, 315
skeleton, xii, 317, 320, 354 stainless steel, 123, 126, 308
skill acquisition, vii, 1, 3, 5, 6, 9, 11, 12, 13, 15, standard deviation, 62, 214
25, 26, 36 standard error, 240, 356
skills, vii, ix, 1, 3, 4, 9, 11, 13, 14, 15, 17, 21, 22, standard model, 352
25, 26, 48, 52, 233, 234, 246, 279 standardized testing, 127
skin, 16, 18, 31, 52, 209, 225, 239, 270, 351, 353, starvation, 183, 195
354, 355, 356, 357, 362, 367, 377, 378, 380, statistical analysis, 83, 85, 86, 334
381 steady state, 236, 256
slipped capital femoral epiphysis, 290, 294 steatorrhea, 139
Slovenia, 281 steel, 123, 126, 308
SMA, 293 stem cells, 312
smooth muscle, 144, 159, 174, 176, 177, 185, stimulus, 215, 217, 218, 219, 220, 221, 226
186, 196, 197, 198, 199, 200, 201 stomach, 138, 139, 141, 144, 145, 157, 160, 161,
smooth muscle cells, 177, 198 165, 168, 170, 175, 176, 177, 182, 188, 190,
smoothing, 357, 377, 380 192, 194, 199, 200, 201, 203, 205
Index 401
strains, 67, 68, 105, 165, 170, 171, 174, 180, 182, thoracic, 139, 158
205, 258 thorax, 16, 374
strategies, ix, 2, 3, 11, 14, 15, 26, 27, 32, 34, 53, threat, ix, 10, 233, 234, 235, 236, 237, 239
136, 154, 228, 233, 234, 236, 237, 244, 245, threats, 10, 237, 239, 241, 243, 247
246, 247 three-dimensional, 54, 105, 106, 128, 145, 168,
stress level, 94, 98, 177, 185, 186 170, 172, 174, 177, 182, 198, 200, 282, 284,
stress-strain curves, 79, 147, 150, 165, 168, 169, 291
175, 181, 183, 185 threshold, 152, 161, 170, 189, 215, 218, 219, 228,
stromal, 59, 60, 61, 63, 65, 74, 80, 82, 83, 84, 96, 239, 240, 241, 242, 243, 344
97, 102, 103, 107, 252, 254, 259, 266 threshold level, 215, 219
subjective experience, 10 thresholds, 44, 140, 182, 204, 216
subluxation, 288, 292, 293 tibia, 130
subjective, xii, 10, 25, 26, 38, 223, 317, 318, 319, time consuming, 90, 376
322, 323, 324, 325, 328, 330, 333, 334, 339, time pressure, 204, 347
submucosa, 142, 143, 144, 149, 153, 166, 167, timing, 17, 18, 21, 119, 208, 212, 215, 224, 270
168, 172, 174, 180, 182 TiO2, 125, 135, 307
substances, 110, 124 tissue engineering, 124
subcutaneous tissue, 211 tissue homeostasis, viii, 137
subcutaneous injection, 161 titania, 128
supramaximal, 214, 215, 216 titanium, 122, 124, 125, 126, 128, 129, 130, 132,
surface tension, 84, 145, 253 133, 134, 135, 136, 306, 310
surgical, 89, 105, 133, 155, 164, 193, 252, 257, titanium dioxide, 307
289, 294, 298, 306 tonic, 153, 174, 177
surgical intervention, 89, 252 tonometry, viii, 57, 58, 89, 102, 103, 107, 110,
stroke, 224, 226, 231, 276 253, 254, 255, 259, 261, 262, 263, 265
survival rate, xi, 305, 306, 311 topographic, viii, 57, 108
susceptibility, 55, 254 torque, ix, 233, 236, 237, 239, 241, 242, 243,
swallowing, 131, 175, 188, 199, 204 244, 245, 248, 296, 299, 302, 352
swelling, 60, 61, 63, 65, 81, 103, 252, 259, 262 total joint replacements, xi, 305
symptoms, viii, 137, 138, 139, 140, 141, 142, trabeculae, 117
145, 152, 154, 157, 158, 159, 160, 161, 182, trabecular bone, 116, 135
202, 222, 288, 289 traction, 188
synchronization, 211 training, x, xi, xii, 3, 6, 7, 11, 12, 13, 15, 18, 20,
synchronous, 214, 216 48, 53, 226, 227, 228, 234, 237, 245, 246, 270,
syndrome, 26, 51, 155, 188, 196, 202, 263 295, 297, 298, 299, 301, 302, 303, 315
synergistic, 235, 298, 299, 300 training programs, 301
synthesis, 52, 150, 155, 195, 248 trajectory, xii, 44, 45, 46, 47, 48, 269, 315
systemic sclerosis, 159, 178, 185, 202 transcranial magnetic stimulation, ix, 207, 224,
226, 230, 231
T
transducer, 71, 166
tensile, 83, 84, 105, 106, 109, 119, 120, 122, 127, transduction, 117, 129, 161, 164
132, 171, 173, 174, 254, 255 transfer, 12, 13, 15, 20, 34, 112, 118, 128, 136
tensile strength, 83, 84, 106, 109, 127, 174, 254, transformation, 379, 382
255 transition, 63, 277
tensile stress, 172 translation, 16, 234, 356, 357, 358, 361, 364, 365,
tensile, 83, 84, 105, 106, 109, 119, 120, 122, 127, 366, 372
132, 171, 173, 174, 254, 255 translational, 372
tensile strength, 83, 84, 106, 109, 127, 174, 254, transmembrane, 209
255 transmission, viii, 111, 112, 122, 131
test-retest reliability, 335 transplantation, 136, 158
therapeutic interventions, x, 234, 237, 244 transport, 50, 61, 141, 158, 165, 166, 172, 175,
therapy, 1, 150, 158, 186, 223, 227, 248, 293, 301 199, 205
thermal expansion, 126 transverse colon, 174, 186
402 Index
trapezius, 224, 228 US Department of Health and Human Services,

trauma, 27, 32, 36, 37, 124, 293 301
trial, 15, 18, 19, 20, 24, 25, 35, 228, 239, 241,
V
263, 273, 278, 303, 321, 330, 332, 366, 367
tribological, 133 valgus, 320
triceps, 270, 271, 272 variability, xii, 26, 190, 214, 216, 224, 229, 230,
triggers, 273 264, 270, 280, 282, 286, 315, 316, 367
trochanter, 282, 283, 286, 287, 353, 354 variables, 21, 30, 236, 248, 271, 316, 318, 324,
tumor, 139, 151, 156 339, 340, 341, 342, 343, 350
Turku, 111 vascular disease, 138
two-dimensional, 291 vastus lateralis, 213
type 1 diabetes, 159, 202 vector, 20, 112, 282, 357, 358, 361, 362, 364,
type 2 diabetes, 157, 160 366, 368, 370, 372
U vessels, 196, 197
veterans, 321, 350
UES, 175, 188, 201 virtual reality, vii, 42, 57
ulcer, 139, 321, 349 viscoelastic properties, 74, 151, 174, 182, 186,
ulceration, 320, 321 203
ulcerative colitis, 186, 200, 202 viscosity, 172, 175, 308
ultrasonography, 164, 170, 190, 197, 198, 200, Visual Analogue Scale (VAS), 318, 319
201
W
ultrasound, 37, 140, 141, 145, 152, 156, 164, 165,
170, 194, 199, 204, 205, 253, 258, 260, 261 walking, 215, 234, 235, 236, 246, 247, 249, 283,
ultrasound biomicroscopy, 260 291, 292, 321, 324, 326, 330, 332, 335, 336,
uncertainty, 358, 366 344, 362, 366, 367, 368, 376, 378, 379, 380
underlying mechanisms, 220, 223, 280, 290 weakness, 13, 234, 243, 244, 245, 369
uniaxial tension, 73, 87, 88 wound healing, 103, 266
uniform, 65, 67, 71, 73, 74, 90, 91, 92, 93, 94, 95,
96, 97, 98, 165, 167, 168, 171, 180, 252, 286, Z
288, 293, 308
urinary, 224, 227 zirconia, 128, 130, 132
zirconium, 128

Biomechanics

Uploaded by

Copyright:

Available Formats

Biomechanics

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Biomechanics

Uploaded by

Copyright:

Available Formats

BIOMECHANICS: THEORY AND APPLICATIONS SERIES

BIOMECHANICS: PRINCIPLES, TRENDS

Biomechanics: Principles, Trends and Applications

BIOMECHANICS: PRINCIPLES, TRENDS

Nova Science Publishers, Inc.

NOTICE TO THE READER

LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA

Available upon request.

ISBN: 978-1-61761-865-9 (Ebook)

Chapter 12 Bone Cell Adhesion: An Important Aspect of Cell Biomechanics in

Biomechanics is the application of mechanical principles (statics, strength of materials

important to understand the pathogenesis to the GI motor-sensory function and dysfunction.

ARTS BIOMECHANICS – AN INFANT SCIENCE:

Gongbing Shan and Peter Visentin

2. PERFORMANCE SKILL ANALYSIS AND ACQUISITION

Documentation pertaining to instrumental performance and dance has a long history. In

Collectively, such sources contain fundamental instruction on music/dance and their

Documentation used a new tone emphasizing the development of technique as a means to

2.2. Challenges Faced by Arts Biomechanics

Notwithstanding efforts to address important artistic concepts through written

Further, success as defined by the performance is a measure which oft-times fails to

2.3. Arts Pedagogy; What Can We Learn from Biomechanics Applications?

In sports, intervention of related disciplines to influence pedagogy and learning is a

1. Characterization of the skill. Characterizing the skill to be acquired involves

1. Cognitive. The beginner focuses on cognitively oriented problems. Performance is

2.4. The Tools of Arts Biomechanics Research

being demanded in human movement and performance research requires synchronized

Figure 12. Rectus abdominis activity at different pitches.

2.5. Samples of Pedagogic Applications

consistencies observed among expert subjects of differing anthropometric proportions can

3. INJURY RISK QUANTIFICATION AND PREVENTION RELATED TO

3.2. Biomechanical Modeling as a Platform to Understand OS

As it is the mechanical work performed by OS patients that obviously causes observed

3.3. The Theory of Biomechanical Modeling as It Applies to OS

3.3.1. The Physics of Mechanical Work

 Work of force F over distance d W=Fd

3.3.2. The Generalized Form Related to Work

W=Fd becomes W=Fvt

W=½p²/m becomes W=½mavt

Thus, the final mathematical representation in all cases is:

MS= cF(t) n dt.

when W, P and t are written as integrals.

3.3.3. Biomechanics and Biological Reactions in the Human Body

VI n dt) m =constant

The differences are that:

 in mechanical PBRs, m may be an integer larger than zero, whereas biomechanical

3.3.4. Characteristics of the PBR related to OS

 OS is the result of the accumulation of micro-trauma

In quantitative terms, the following fundamental premises may be used:

1. Since damage (D) = MicroDamage  , it is obvious to put damage

Damage (slow) Repair (slow) Cummulative damage (slow)

AMT = (I)ndt  Rdt

3.4. Current OS Studies Related to Music Performance

3.5. Future Directions

Stage 1. Determination of Realistic Loading Conditions

Stage 3. Determination of Accumulated Micro-Trauma

4. APPLICATIONS OF MOVEMENT SCIENCE TECHNOLOGIES IN THE

Figure 20. Diskobolos of Myron (circa. 460 BC).

Bernstein, who developed precise kinematic procedures using cyclography, a technique

Figure 21. Multi-camera stop action recording of a pole-vaulter, by Muybridge.