See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/49731928

Basic functional electrical stimulation (FES) of extremities: An engineer's

view

Article in Technology and health care: official journal of the European Society for Engineering and Medicine · January 2010
DOI: 10.3233/THC-2010-0588 · Source: PubMed

CITATIONS READS

8 2,832

2 authors:

Tadej Bajd Marko Munih

Research Centre of the Slovenian Academy of Sciences and Arts University of Ljubljana
236 PUBLICATIONS 2,816 CITATIONS 252 PUBLICATIONS 2,462 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

CYBERLEGs View project

CYBERLEGs View project

All content following this page was uploaded by Marko Munih on 03 March 2014.

The user has requested enhancement of the downloaded file.

 VI.2: Basic Functional Electrical
Stimulation(FES) of Extremites
– an Engineer's View
Tadej BAJD and Marko MUNIH
Faculty of Electrical Engineering
University of Ljubljana
Slovenia

Abstract.

Keywords.

Introduction: Brief History of FES

Medical applications of the discharge of electric fish for the treatment of headache and
arthritis were first recorded by Scribonius Largus in the second century AD. However,
the first electric devices making use of static electricity were developed in the
eighteenth century. These were based on the principle of using friction of insulators,
such as glass, to develop high voltages due to charge separation. The ninteenth century
brought a more practical stimulator with an induction coil. The so-called 'Faradic
stimulator' consisted of a mechanical interrupter in series with a battery and the primary
winding of a transformer. The output from the secondary winding of the transformer
was a series of pulses which were similar to the stimuli of the present day electrical
stimulators. The Faradic stimulator was the first device that could produce controlled
and sustained muscle contractions. The output frequency could be adjusted to easily
exceed the stimulation rate of 100 pulses per second and the output level was also
controllable. Finally, the twentieth century brought transistors, integrated circuits, and
microprocessors permitting sophisticated electronic circuits to be incorporated into
very small devices.

1. FES Parameters

Functional electrical stimulation (FES) is a rehabilitation technology that uses electrical

currents applied to peripheral nerves. When a stimulating current is applied to the
electrodes placed on the skin overlying sensory-motor structures, an electric field is
established between two electrodes and ions will create a current in the tissue (Figure
1).
 Figure 1. Electric field between a positive and negative electrode.

The ionic flow across the nerve influences transmembrane potential and can
generate an action potential. The action potential propagates along the nerve causing
contraction of a paralyzed muscle. In this way, FES provides restoration of movement
or function, such as standing or walking by a person with a spinal cord injury.
FES is performed in a series of rectangular monophasic or biphasic (symmetrical
or asymmetrical) electric pulses described by the following parameters: amplitude or
intensity of pulses, frequency or pulse repetition rate, duration of single pulse, and
duration of a pulse train. In most cases of surface FES applications, periodic
monophasic or unidirectional pulses are used. Biphasic or bidirectional pulses prevent a
slow deterioration of the electrodes, while the chemical conditions on the skin and in
the muscular tissue remain unchanged.

1.1. Stimulator Circuit

With respect to the stimulator output circuit, electric pulses are either voltage- or
current- controlled. Stimulators providing a constant output voltage can maintain the
voltage desired, irrespective of resistance changes in the stimulated muscle tissue.
Stimulators with current output stages make constant current pulses possible. An
important difference between the two stimulator types becomes evident in the case of
an improper contact between the electrode and the skin. In the case of a constant
current stimulator, a smaller effective electrode surface results in greater current
density, which can cause skin burns. With a constant voltage source, the resistance
increases due to insufficient contact, which results in a decrease of current and,
consequently, of the muscle response, but causing no skin damage.

1.2. Muscle Recruitment

A muscle recruitment curve represents the dependence of isometric (measurement

performed at a constant muscle length) joint torque upon the FES amplitude or pulse
duration (Figure 2). The joint torque is not linearly dependent upon the stimulation
intensity. Two nonlinearities occur - threshold and saturation. The increase in joint
torque due to an increasing amplitude of electrical stimulation occurs as a result of
activating new fibers in a nerve bundle laying in an electric field between the electrodes.
The main reason why all nerve fibers do not react to the same stimulation amplitude is
found in the differences in the stimulation threshold and various distances from the
stimulation electrodes. First, the fibers closest to the electrodes are stimulated. In
addition, the fibers with a greater diameter respond earlier. Beyond a certain
stimulation intensity, the force of contraction no longer increases. At such a stimulation
amplitude, all nerve fibers are excited, and a further increase of the stimulus does not
increase contraction. In surface FES of knee extensors for example, the values of the
stimulation threshold range between 20 and 60 V, while the saturation value is between
100 and 150 V.

Figure 2. Muscle recruitment curves assessed in 10 paraplegic subjects after a FES restrenghtening
programme.

A single stimulation pulse provokes only a short-lived muscle twitch of no more

than 0.2 s. If electrical stimuli are repeated every second, a twitch occurs every second,
between which the muscle relaxes. If the frequency of stimulation pulses increases up
to 10 pulses per second (10Hz), between two twitches there is no time left for muscle
relaxation. When measuring isometric contraction, we get twitching responses. This
twitching is considerably reduced at stimulation frequencies between 15 and 20 Hz
(Figure 3). At higher frequencies, the response is already smooth: this is known as
tetanic contraction. The frequency at which the tetanic contraction occurs is called
fusion frequency. It is not the same for all muscles and depends on properties of muscle
fibers. Changes in stimulation frequencies also affect the intensity of the response. As
regards the response intensity, slight losses are observed at lower stimulation
frequencies. On the whole, the changing of frequency between 40 and 100 Hz causes
small differences in the isometric torque as measured in the joint.

Figure 3. Influence of stimulation frequency on muscle response.

1.3. Muscle Fatigue

A low stimulation frequency results in less pronounced fatigue of the neuromuscular

system. An electrically stimulated muscle fatigues more quickly than in the case of
voluntary contraction. The main reason is the reversed recruitment order. It is not too
serious an oversimplification to think of human muscle as a combination of slow fibers,
capable of sustaining low levels of contractile activity without fatigue for prolonged
periods, and fast fibers, capable of developing large forces, but fatiguing so rapidly that
they can be used only in intermittent activities. Some muscles are predominantly made
up of slow fibers, some predominantly of fast, and some of a given mixture of the two.
All muscle fibers innervated by the same motoneuron have been found to be of the
same type. Motoneurons innervating predominatly slow muscles have axons of small
diameter, and those supplying fast muscles have axons of a larger diameter. In a
voluntary contraction of a normally innervated muscle, the slow fibers are recruited
first, and as increased muscle force is required, the fast fibers are recruited. Slow fibers
are, therefore, activated frequently, while fast fibers are employed only infrequently,
during a burst of intense activity. When applying electrical stimulation, fibers with a
greater diameter respond earlier. This are motoneurons innervating fast muscles. The
normal order of recruitment order is, therefore, reversed resulting in an increased
fatiguing of electrically stimulated muscle. In addition, by electrical stimulation, the
same nerve fibers are stimulated all the time, whereas with a healthy muscle the work is
divided among different motor units of the same muscle. Also, due to relatively high
stimulation frequency, the transmitter in the neuromuscular junctions is being
exhausted, so the muscle stimulated soon shows signs of fatigue.

1.4. Duration of Stimulus

In a fashion similar to amplitude, pulse duration also exerts a direct effect upon the
intensity of contraction. Here again, this is determined by threshold value and response
saturation. When applying surface stimulation electrodes, the accompanying unpleasant
sensation (when preserved) or even skin damage is due mainly to an excessively long
duration of a stimulus, therefore short durations are used (0.1 to 0.5 ms), while the
force of a paralyzed extremity is controlled by increasing the stimulus amplitude.
Changing the pulse duration has little or no effect on stimulated muscle fatigue.
Functional movement of a paralyzed extremity cannot be obtained by a single
electric stimulus, but requires a series of stimuli of a certain duration, following one
another at an appropriate frequency. Such a series of stimuli is called a stimulation
pulse train. In FES training of atrophied muscles, a stimulation pulse train is followed
by a pause, and then by another stimulation train. The relationship of train duration and
pause is often called the duty cycle and exerts an influence upon the fatigue of a
stimulated muscle.

2. FES Electrodes

2.1. Anode & Cathode

A surface stimulation electrode is a terminal through which electrical current passes

into the underlying tissue. At the electrode-tissue interface, a conversion occurs
between the current of electrons driven through the wires coupled to the stimulator and
the current of ions in the tissue. An electrode is usually made of metal. However, it
may be made of a nonmetal, commonly carbon. The electrode through which current
passes from the metallic or nonmetallic conductor to the tissue is called the anode and
that through which current passes from the tissue to the conductor is the cathode. In
electrical circuits, the current flows from the terminal at higher electrical potential to
the terminal at lower electrical potential. In this way the anode is the positive electrode
and the cathode the negative electrode.

2.2. Unipolar & Bipolar

Besides distinguishing between positive and negative electrodes, we also speak about
unipolar and bipolar electrical stimulation techniques. With unipolar stimulation, one
electrode is often considerably smaller than the other, whereas the electrodes used in
bipolar stimulation both have the same size. In unipolar stimulation the smaller
electrode is negative and is also called an active electrode due to the fact that in its
vicinity there occurs depolarization of the membrane of nerve fibers. In motor nerve
stimulation, the active electrode is positioned as closely to the motor point of the
muscle as possible. The motor point is a site on the skin, where the amplitude of the
stimulus required to fully activate the muscle is at a minimum, and where all of the
motor nerve fibers are closest to the stimulating electrode. In multichannel electrical
stimulation systems, it is possible to have a single anode and several independent
cathodes or to have anodes and cathodes that are galvanically separated.
Let us examine four properties of surface stimulation electrodes and electrodes
positioning, which influence the effectiveness of electrical stimulation: electrode size,
polarity of electrodes, resistance, and distance between the electrodes.
(i) Electrode size: Electrical stimulation is applied to a nerve fiber, since muscle
fibers have a considerably higher stimulation threshold. Thus we can say that larger
electrodes are used to stimulate the nerve endings spreading all over the underlying
tissue, whereas smaller electrodes are applied to influence the nerve when the latter
come closer to the skin. Using larger electrodes, stronger contraction is obtained along
with a reduced current density and a less pronounced unpleasant sensation on the skin.
However, large electrodes permit no selective choice of a desired movement of the
stimulated paralyzed extremity. The active areas of electrodes range between 2 cm2 and
50 cm2. Electrodes of 2 cm2 to 4 cm2 are used to stimulate the nerves near the surface,
those of about 8 cm2 for the stimulation of smaller muscles, while electrodes of 25 cm2
or more are used in case of larger muscles.
(ii) Polarity:A positive and a negative electrode are placed along the muscle to be
stimulated. Considering their polarity, the electrodes are positioned so as to provoke an
optimal movement from the functional point of view. Stronger movement is usually
obtained by placing the positive electrode distally.
(iii) Resistance:It is desirable that the resistance should be as low as possible in
order to avoid energy losses before the stimulation has reached the neuromuscular
tissue. The impedance between the electrode and the skin is frequency dependent. The
DC (or low frequency) impedance tends to be several orders of magnitude larger than
the impedance at higher frequencies. Nominal values of 2 kΩ are encountered. Contact
conduction is increased by moistening the electrodes with water or special conductive
electrode gels. Adipose tissue offers high resistance to electrical currents and so higher
stimulation amplitudes must be used, causing pain in the skin as a side effect. Bones,
too, are very bad conductors of electric current; electrical stimulation cannot reach
muscles which are behind them.
(iv) Distance between electrodes:The greatest current density appears at the skin-
electrode contact and tends to decrease with distance from the electrodes as the flow
spreads out over a larger area. Closely spaced, small electrodes generally make the
effective area of stimulation rather superficial due to the lower impedance of the
current path through proximal tissue. Deeper tissues will be stimulated by using a
greater distance between the electrodes. Increasing the electrode separation leads in
general to an increase of the maximal achievable force. If the skin between the
electrodes is moist, this causes the current between the electrodes to flow to the skin
which results in a burning sensation and a slight or no muscle contraction at all.

2.3. Electrode design criteria

The design criteria for surface stimulation electrodes are: physical comfort to the skin,
electrical surface area greater than four square centimeters to prevent skin irritation, use
of hypo-allergenic materials, flexibilty to follow body surface, ease of attachement and
ability to remain in position for a duration of at least one active day, reusable, low cost,
reliable means of connection to stimulator, resistant to medical solvents and electrode
gels, low and stable electrical resistance.
The simplest surface electrodes consists of a metal plate or metal wire mesh coated
with fabric or sponge. Common materials used are stainless steel, silver-silver chloride,
platinum, or gold. For safety purposes, the upper part of the electrode is covered with a
non-conductive material. The electrode is applied after having been moistened with
water. Such electrodes are usually fixed on the extremity by means of velcro or elastic
bands. With the development of longer term surface electrodes, these electrodes are
often used for site evaluation either in pain treatment or stimulation provoking muscle
contraction. Small, button-shaped electrodes of similar design are highly suitable for
the stimulation of a single nerve. Here, the metal plate is coated with several layers of
gauze so that the electrode might retain the moisture required for as long as possible.
Surface electrodes made of silicone impregnated with carbon are applied to the
skin surface with conductive gels and held in place with adhesive patches. A too thinly
or unevenly spread gel increases current density at certain points, thereby bringing
about a danger of burns. Electrodes may be left on the skin for several days a time.
Another important property of electrodes made of conductive rubber is their flexibility
making them adaptable to any part of the body. These electrodes can be shaped by
cutting, so as to adapt them as much as possible to a proper stimulation site.
Conductive adhesive gel electrodes provide self-adhesion when applied (Figure 4).
Karaya gum was the first of these adhesive gels. Later a variety of electrically
conductive polymers were developed, enabling good contact with the irregularities of
the skin surface and a more uniform current density at the electrode-skin interface.
These electrodes can be used for extended periods with minimal skin irritation. An
important breakthrough in electrode design was made by Axelgaard. This is a flexible
electrode which adheres well to the patient's skin, is easily removed therefrom, and is
able to move with the patient's skin ensuring proper placement of the electrode. The
electrode is in the form of a knit conductive fabric. Conductive fibers include a blend of
stainless steel and polyester. The fabric can be stretched up to about 20%. A conductive
adhesive fills interstitial areas of the knit fabric and adheres the electrode to the
patient's skin. A non-conductive sheet on the other side of the knit fabric prevents
undesired electrical contacts.
 Figure 4. A conductive adhesive gel electrode with a portion of nonconductive sheet peeled back showing
the knit fabric.

2.4. Problems with electrodes

When improperly handled, electrodes can damage the skin in the contact area. Burns
typically occur underneath the anode, but not the cathode, when using identical surface
electrodes. Another problem resides in a precise electrodes positioning along a muscle.
Sometimes a displacement of a few millimeters completely changes the muscle
response. This happens when a selected nerve (e.g. peroneal nerve) ought to be
stimulated by surface electrodes. Surface electrodes may excite pain receptors in the
skin, although patients sensibility may be reduced to such an extent that the sensation
of pain is not critical. Another problem is undesired motion of the skin with respect to
the neuromuscular tissue. Even though an electrode seemingly occupies the same place
all the time, its distance from the nerve is not constant. This is one of the reasons why
the movements caused by electrical stimulation cannot be easily repeated. Another
limitation is that small muscles usually cannot be selectively activated and deep
muscles cannot be stimulated without first exciting the superficial muscles. Relatively
high voltages, sometimes in excess of 100V, between electrode pairs cause hazards for
the patients and the personnel that treat them. Finally, the applicability of the surface
stimulation electrodes depends on fixation problems. Stretchable garments with
electrodes already mounted in appropriate locations have been developed by several
manufacturers to simplify the application of electrodes to the skin surface. In the case
of lower limb stimulation, fixation problems can be overcome by specially designed
trousers carrying stimulation electrodes and cables. Such stimulation equipment is
comfortable and easy to handle. In the non-invasive, upper limb neuroprosthesis, the
surface stimulation electrodes were built into an elegant, self-aligning, and flexible
splint. The splint provides additional fixation of the wrist joint and allows the entire
electrode array to be positioned within a few seconds.
It is not difficult to realize that most of the inconveniences of surface stimulation
electrodes can be overcome by the use of implanted electrodes. Nevertheless, because
of their simple non-invasive application, surface electrodes will remain of use in
therapeutic treatments.

3. Electrical stimulators

Electrical stimulators comprise an input circuit, pulse generator, output stage, and
power supply (Figure 5). The input into the electrical stimulator is represented by a
control signal automatically switching electrical stimulation pulses on or off or may be
under the patient's voluntary control. The output electric pulse current is led, via
electrodes, to the selected stimulation site. The type of stimulation pulse is determined
by a pulse generator. The output pulses provided by the circuit are considerably lower
than the surface stimulation pulses required for a functional movement of an extremity
but are of an appropriate frequency and duration. The output stage ensures energy for
the electrical stimulation of paralyzed muscle in either constant current or the constant
voltage outputs. In case of current stimulation, the internal resistance of the end stage is
considerably higher than the tissue resistance between the electrodes. The current
source of stimulation pulses provides a constant current irrespective of the resistance of
the skin and the tissue between the electrodes. In the case of constant voltage output
stages, skin resistance is lower than that between the electrodes, and the stimulator
provides a constant voltage independent of the skin and tissue resistance. A power
supply provides the energy necessary for the operation of particular electronic circuits
(low voltage) and the electrical stimulation itself (high voltage). Electrical stimulators
are usually battery powered with stimulation pulses with an amplitude of more than
100 V. A high voltage for the output stage is obtained from low battery voltage by
means of a voltage convertor.
The development of electronics made it possible for Wladimir Liberson to develop
a stimulator to preventing foot-drop in hemiplegic patients. It was triggered by a heel
switch in the shoe of the affected leg. Stimulation electrodes were positioned above the
peroneal nerve in the popliteal fossa, behind the knee joint. Each time the patient raised
the heel, the heel switch triggered the stimulator, causing the nerve to cause the
extensor group of muscles to contract and dorsiflex the ankle and so lift the foot.
Advanced versions of these dropped foot stimulators were developed in Ljubljana,
Slovenia and Salisbury, UK. Both stimulators were applied to a large group of stroke
patients. In the WalkAide peroneal nerve stimulator, a tilt sensor triggers ankle
dorsiflexion.

Input Pulse Output

Circuit Generator Stage

Control Electrodes

Power
Supply

Figure 5. Block scheme of an electrical stimulator.

A minimum of four channels of FES were used for synthesis of a simple reciprocal
gait pattern in completely paralyzed paraplegic subjects. During reciprocal walking, the
stimulator must be controlled through three different phases of walking: right swing
phase, double stance phase, and left swing phase. This is achieved by two hand-push
buttons built into the handle of the walker or crutches. When niether of the push
buttons is pressed, both knee extensors are stimulated providing support to the body.
On pressing the push button in the right hand, the peroneal nerve is stimulated in the
right leg eliciting the flexion reflex presented by simultaneous hip and knee flexion and
 ankle dorsiflexion. The same is true for the left leg. The Ljubljana FES walking system
consists of two small two-channel stimulators attached to each leg. Only three
electrodes are applied to a single leg in order to produce knee extension and flexion
responses. As both activities never occur simultaneously, the distal electrode placed
over the knee extensors represents the common electrode for both stimulation channels.
Using the same principles as the Ljubljana FES system, and adding two channels of
stimulation to both hip extensors, the FDA-approved Parastep surface stimulation
system was developed. A new multipurpose programmable transcutaneous electric
stimulator, Compex Motion, was developed to allow users to design various custom-
made neuroprostheses, neurological assessment devices, muscle exercise systems, and
experimental setups for physiological studies. The Compex Motion stimulator can
generate any arbitrary stimulation sequence, which can be controlled in real-time using
any external sensor.
Despite the fact that correct application of a stimulator implies no danger either for
the patient or therapist, we wish to emphasise some important safety points. When
buying a new electrical stimulator, be sure that it carries the CE mark. Every stimulator
has its own characteristics which must be indicated in the instructions of use, an
obligatory accompaniment of any commercially-available stimulator. Special attention
should be paid to stimulation parameters: pulse duration, frequency, and maximal
current or voltage. Also of utmost importance is the information on the proper use of
surface electrodes. The use of electrical stimulators might be dangereous in case of
patients with an implanted pacemaker. Simultaneous use of electrical stimulators and
high frequency surgical devices is prohibited. Electrical stimulators might not work
properly in close proximity to microwave devices. Transthoracic positioning of FES
electrodes may cause fibrillation of the heart. The stimulator should not be switched on
in case of short-circuit of the electrodes. It is true, that electrical stimulators are made
in such a way that a short-circuit of several minutes does not damage them. However,
one should not take unfair advantage of this property.

Further Reading

[1] L.A. Benton, L.L. Baker, B.R. Bowman, R. Waters, Functional electrical stimulation – A practical
guide, Rancho Los Amigos Hospital, 1980
[2] L. Vodovnik, T. Bajd, A. Kralj, F. Gračanin, and P. Strojnik, Functional electrical stimulation for
control of locomotor systems, CRC Critical Rev. Bioeng. 6 (1981), 63-131
[3] A.R. Kralj and T. Bajd, Functional electrical stimulation: Standing and walking after spinal cord
injury, CRC Press Inc., 1989
[4] L.S. Illis (ed.), Neurological Rehabilitation, Blackwell Scientific Publications, 1994
[5] D. Popović and T. Sinkjaer, Control of movement for the physically disabled, Springer, 2000
[6] P.J. Rosch and M. Markov (eds.), Bioelectromagnetic Medicine, Marcel Dekker Inc., 2004
[7] K.W. Horch, GS Dhillon (eds.), Neuroprosthetics – Theory and Practice, World Scientific, 2004
[8] J.D. Bronzino (ed.), The Biomedical Engineering Handbook, CRC Press and IEEE Press, 2000
[9] M. Akay (ed.), Wiley Encyclopedia of Biomedical Engineering, John Wiley&Sons, 2006
[10] L. R. Sheffler, J Chae, Neromuscular electrical stimulation in neurorehabilitation, Muscle Nerve 35
(2007), 562-590

View publication stats