biosensors
Article
Computation of Vascular Parameters: Implementing
Methodology and Performance Analysis
Mohamed Yacin Sikkandar 1, * , Sridharan Padmanabhan 2 , Bobby Mohan 2 , Ibrahim AlMohimeed 1 ,
Ahmad Alassaf 1 , Shady A. Alshewaier 3 , Ali Abdullah Almukil 1 and Sabarunisha Begum 4
1
2
3
4
*
Citation: Sikkandar, M.Y.;
Padmanabhan, S.; Mohan, B.;
AlMohimeed, I.; Alassaf, A.;
Department of Medical Equipment Technology, College of Applied Medical Sciences, Majmaah University,
Al Majmaah 11952, Saudi Arabia
Department of Biomedical Engineering, Rajalakshmi Engineering College, Chennai 602105, India
Department of Physical Therapy, College of Applied Medical Sciences, Majmaah University,
Al Majmaah 11952, Saudi Arabia
Department of Biotechnology, P.S.R. Engineering College, Sivakasi 626140, India
Correspondence: m.sikkandar@mu.edu.sa
Abstract: This paper presents the feasibility of automated and accurate in vivo measurements of
vascular parameters using an ultrasound sensor. The continuous and non-invasive monitoring of
certain parameters, such as pulse wave velocity (PWV), blood pressure (BP), arterial compliance
(AC), and stiffness index (SI), is crucial for assessing cardiovascular disorders during surgeries and
follow-up procedures. Traditional methods, including cuff-based or invasive catheter techniques,
serve as the gold standard for measuring BP, which is then manually used to calculate AC and SI
through imaging algorithms. In this context, the Continuous and Non-Invasive Vascular Stiffness
and Arterial Compliance Screener (CaNVAS) is developed to provide continuous and non-invasive
measurements of these parameters using an ultrasound sensor. By driving 5 MHz (ranging from 2.2
to 10 MHz) acoustic waves through the arterial walls, capturing the reflected echoes, and employing
pre-processing techniques, the frequency shift is utilized to calculate PWV. It is observed that PWV
measured by CaNVAS correlates exponentially with BP values obtained from the sphygmomanometer (BPMR-120), enabling the computation of instantaneous BP values. The proposed device is
validated through measurements conducted on 250 subjects under pre- and post-exercise conditions,
demonstrating an accuracy of 95% and an average coefficient of variation of 12.5%. This validates the
reliability and precision of CaNVAS in assessing vascular parameters.
Alshewaier, S.A.; Almukil, A.A.;
Begum, S. Computation of Vascular
Keywords: arterial compliance; stiffness index; pulse wave velocity; blood pressure; acoustic waves
Parameters: Implementing
Methodology and Performance
Analysis. Biosensors 2023, 13, 757.
https://doi.org/10.3390/
bios13080757
Received: 21 May 2023
Revised: 20 July 2023
Accepted: 22 July 2023
Published: 25 July 2023
Copyright: © 2023 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
1. Introduction
According to the World Health Organization (WHO), cardiovascular diseases (CVD)
are the leading cause of global mortality, affecting approximately 17.9 million individuals
annually. Hypertension has impacted an estimated 1.13 billion people worldwide, with
the majority residing in low- and middle-income countries. Those with hypertension are
at a higher risk of developing severe medical complications, including brain, heart, and
kidney diseases, among others. Disturbingly, reports indicate that 1 in 4 men and 1 in
5 women suffer from hypertension, making it a significant contributor to premature mortality on a global scale. Physiologically, pulse wave velocity (PWV), blood pressure (BP),
arterial compliance (AC), and stiffness index (SI) are closely related vascular parameters
and play an important role in maintaining blood flow to all vital organs [1–3]. Studies
have demonstrated a significant correlation between these four parameters and cardiovascular diseases [4–6]. Measuring and monitoring these parameters non-invasively and
continuously is a main concern for clinicians, and there is a need for an automated device,
specifically throughout prolonged medical procedures like dialysis and surgeries. The
first conference of consensus on arterial stiffness (AS) held in June 2000 (Paris, France)
Biosensors 2023, 13, 757. https://doi.org/10.3390/bios13080757
https://www.mdpi.com/journal/biosensors
Biosensors 2023, 13, 757
2 of 13
summarized the existing devices for measuring arterial compliance depending on principles like pulse transmit time, arterial pressure pulse, and vessel diameter, but these are
targeted only to measure arterial compliance [7]. Manual measurements of blood pressure
based on the gold standard sphygmomanometer demands focused attention and can lead
to human errors if operated by an unskilled individual. A non-invasive measurement
of AS and AC using traditional oscillometric blood pressure measurement was carried
out by Hidehiko et al. [8]. A calibration-free Photoplethysmography (PPG) waveform
analysis and a biometric-based technique were used to measure BP [9]. In addition to the
above theoretical procedures, there are many patented devices available to monitor BP
non-invasively and continuously using cuff-based pressure sensors [10–13]. Hoctor et al.
developed an ultrasonic-based method and apparatus to measure BP through a continuous
and non-invasive method by capturing images using ultrasound to derive blood pressure
values [14]. Recently, Yamil Kuri developed a system and method to determine AC and SI.
This system was configured to calculate BP based on calculating the blood flow velocity
and subsequently derived AC and SI [15,16]. Polanczyk A et al. have conducted hemodynamic studies on a three-dimensional reconstruction of the blood vessel and describe its
biomechanical properties [17–19].
All these devices monitor PWV, BP, AC, and SI parameters individually or collectively,
either by causing discomfort to the subject with continuous measurements or image-guided
ultrasound modalities with complex algorithms. Also, these devices require a large setup for monitoring in intensive clinical environments, which makes it more complex for
implementation in basic healthcare units in remote areas. Computer- and image-assisted
methods of diagnosis have high accuracy, but repeatability remains a big area to be worked
on. Many researchers have discussed carotid stiffness indices, which depend on the relationship between the pressure and the diameter of arterial distension from the diastolic to
the systolic phase, giving deep insight into the relationship between geometrical parameters of the blood vessel and clinical values [13,20–29]. The in vivo non-invasive technique
that employed Young’s modulus estimated from the regional stress–strain relationship
conveyed wisdom learning on arterial wall properties [30]. Few studies have worked on the
in vivo Young’s modulus measurement of the pressure–strain relationship on the carotid
artery or based on the slope of the stress–strain relationship at end-diastole and end-systole.
Young’s modulus of the arteries, which is a relationship between blood pressure and arterial
diameter, is found to indicate profound variations for different individuals [31–36]. Existing
ultrasound imaging has been widely employed in the imaging of soft tissues, specifically
blood vessels, for diagnosing cardiovascular disorders. The movement of the arterial wall
can be measured using the M-Mode ultrasound method of estimation or using RF signals.
These imaging methods have their targets specifically on brachial, femoral, and carotid
arteries and abdominal aorta [37]. Furthermore, 1D cross-correlation methods have been
employed to estimate carotid distention waveform and to visualize its structure [38].
A survey of the literature indicates that existing modalities employed to find blood
pressure and arterial parameters have many impediments. Frequent usage of cuff-based
devices may lead to damage of the blood vessels, and a skilled technician is required to
detect the Korotkoff sound to evaluate BP. Employing separate modalities like imaging
algorithms for finding arterial parameters and invasive catheter methods to detect continuous blood pressure leads to an increase in complexity and a potential clinical hazard that is
unavoidable during long-term surgical procedures.
Clustering all of these important parameters into a single continuous non-invasive
device without a complex imaging algorithm would benefit clinical specialists to successfully carry out prolonged and lifesaving surgeries. In this context, the aim of this research
is to develop a device that will measure said parameters continuously and non-invasively
using an ultrasound sensor. This paper provides additional confirmation of the potential clinical application of a Continuous and Non-Invasive Vascular Stiffness and Arterial
Compliance Screener (CaNVAS) device by demonstrating the feasibility of automated and
precise in vivo measurements of these parameters using a straightforward and innovative
Biosensors 2023, 13, 757
3 of 13
methodology. This validation paves the way for the device’s potential use in clinical settings
in the near future.
2. Materials and Methods
This device focuses on deriving the blood pressure value from PWV; in addition, AC
and SI are computed, making it a suitable alternative for existing modalities. Nonionizing
ultrasound waves are passed through the blood vessel to obtain the reflected echoes to
evaluate the PWV using the Doppler shift. The relationship between PWV and BP is established, which helps in the automated estimation of BP from PWV. AC and SI are formulated
by extracting the derived systolic and diastolic diameter and blood pressure. To validate
the entire experimental set-up in its working, two checkpoints with the artery diameter
value as an indicator are programmed to indicate any deviation on device placement in the
region of interest. A vessel finder is utilized to trace the artery position or to correct the
CaNVAS position whenever needed. CaNVAS is developed to monitor vascular parameters
under a single roof in a feasible manner without involving any complex algorithm, making
it extendable to bett employed as a point-of-care device for CVD patients.
tt
Figure 1A illustrates the comprehensive architecture of CaNVAS, which consists of
various modules essential for measuring vascular parameters. These modules include:
2.1. CaNVAS: System Architecture
(i)
A sensor positioning module equipped with infrared illumination to mark the artery’s
tt
position.
(ii) An embedded ultrasound sensor module housing a piezoelectric crystal oscillator
with an appropriate frequency.
(iii) A transmitter unit, receiver unit, and pre-processing module encompassing an acquisition unit, operational amplifier, and high-pass filter.
(iv) A microprocessor and signal processing module responsible for extracting vascular
parameters.
(v) A ttdisplay module featuring a manual setting mode and an output display unit.
Figure 1. (A). CaNVAS system architecture (B). CaNVAS device sketch.
Biosensors 2023, 13, 757
4 of 13
It is important to note that the term “vascular parameters” encompasses systolic blood
pressure (Ps ), diastolic blood pressure (Pd ), pulse wave velocity (PWV) (which includes
peak systolic velocity (Vs ) and end diastolic velocity (Vd )), arterial compliance (AC), and
stiffness index (SI).
2.1.1. Sensor Positioning Module
A sensor positioning module is provided with an artery tracing circuit to support
the non-invasive device in marking the correct position of the target artery on the skin’s
surface, thereby illuminating the infrared source at a wavelength of 620 nm to visualize the
target artery. Figure 1B shows the device sketch in which the sensor is positioned on the
surface of the wrist underneath the display panel. The device includes an ultrasound sensor
transmitter that is embedded with a piezoelectric crystal oscillator and a driver circuit. This
transmitter emits a continuous acoustic wave at a frequency of Ft . At the other end, the
receiver module picks up the reflected acoustic wave FO , which exhibits a distinct frequency
compared to the transmitted frequency. This frequency difference between the transmitter
and receiver modules is known as the Doppler shift (FS ). The device transmits acoustic
waves at a predetermined frequency across the blood vessel interface to capture reflected
waves from the vessel’s surface, causing a frequency shift in the receiver module. Both the
transmitter and receiver modules are positioned adjacent to each other on the skin’s surface,
aligned with the direction of blood flow towards the receiver module. A crystal oscillator
embedded in the sensor activates the ultrasonic transmitter module, generating waves at
a specific frequency that interact with the blood vessel. After being absorbed, the waves
are reflected back and detected through the oscillator crystal in the receiver module. The
programmed microprocessor controls the crystal oscillator, producing specific frequencies
tailored to the targeted arterial site. More specifically, the transmitting frequency Ft is
generated in a range between 1.5 and 3.5 MHz for the brachial artery, 4 and 7 MHz for the
radial artery, and 7 and 10 MHz for the carotid artery (with an ultrasound sensor with 98%
accuracy with a resonant frequency variation of ±5%) to measure the vascular parameters.
The range of frequencies is provided to allow for the selection of a specific frequency that
is appropriate for the individual subject being examined. The CaNVAS device is tailored
for three specific arteries: radial, brachial, and carotid. Each artery is outfitted with its
ultrasound sensor and a corresponding algorithm, guaranteeing the emission of acoustic
waves at precise frequencies targeted to the specific arterial site.
2.1.2. Pre-Processing Module
The driver circuit is designed in such a way that it guides to transmit the ultrasound acoustic wave at the appropriate frequency towards the target arterial site; the
pre-processing module comprises an operational amplifier and a high-pass filter, wherein
the received acoustic frequency FO is converted to an electrical signal by a receiver crystal.
That electrical signal is amplified by an operational amplifier with a bandwidth greater
than 200 MHz and a slew rate of, at a minimum, 4100 Vµs−1 . The operational amplifier
can amplify the electrical signal by a gain value beyond 25. Furthermore, the receiver
module is coupled to a high-pass network with a cut-off frequency of 3 MHz to filter the
fluctuations on the amplified electrical signal, allowing the sharp amplified electrical signal
to microprocessor. This module is programmed with a structural algorithm to extract
vascular parameters from the detected reflected acoustic waves. Two checkpoints, Young’s
modulus validation to compute the vessel diameter and a diameter-varying function to
correct the device position, are included within the algorithm to verify the results obtained
simultaneously. The validation unit is also incorporated to validate the pressure value from
CaNVAS by the operator. A display module continuously provides the data to the user or
attender. The data include the instantaneous systolic blood pressure PS, the instantaneous
diastolic blood pressure Pd , the pulse wave velocity, which comprises vs. and Vd , the
arterial compliance, the stiffness index of the user, the artery position marking, and the
device validation report.
Biosensors 2023, 13, 757
tt
5 of 13
ff
2.1.3. Computation of Vascular Parameters
A unique methodology is carried out in CaNVAS to determine the constants α and β,
which are fed into the predetermined Equation (1), expressing the nonlinear relationship
α
between the pulse wave velocity and the pressure.
β
P = αV2 + β (kPa)
(1)
P = αV + β
The Doppler shift Fs , which is computed using the received echo signal from the
receiver, is taken as an input for Equation (2) to derive the PWV [37] measure in ms−1 .
−
C × Fs
PWV
= =
(ms−−1 )
PWV
2cosθ × Ft
(2)
−1
C is the speed of the acoustic wave in the artery, which is 1540 ms−
; FS is the doppler shift,
because the sensors are placed adjacently; andcosθ
cosθ is taken as the integrated angle value
◦
◦
cos30°, which is 0.5. Ft is a transmitter
tt frequency, which is selected in
between cos0
cos0° and cos30
the specific range from about 1.5 MHz to 10 MHz based on the target artery, wherein the
target arteries are characterized as the radial artery, the brachial artery, and the carotid
artery. The instantaneous pulse wave velocity obtained from the subject is used to extract
the peak systolic velocity (Vs ) and the end diastolic velocity (Vd ). The minimum and
maximum PWV in one cycle duration is determined, of which 70% of the maximum value
is taken as vs. and 10% of it is taken as Vd . The reference blood pressure (Pr ) parameter
from a multitude of 995 random subjects, which included divergent age groups from 21
to 55 (38 ± 17) with various physiological condition (diabetic, high BP, and acute cardiac
issues) was measured by means of an external blood pressure measuring device, Diamond
Deluxe BPMR-120. Furthermore, the reference pulse wave velocity (Vr ) from the same
random subjects using an initial CaNVAS experimental set-up is programmed to compute
only the PWV (the procedures were explained to all the subjects included for study and
non-objection consent forms were collected before recording the data). These reference
pressure values (Pr ) and pulse wave velocity values are plotted
on a graph, as depicted in
tt
Figure 2, to find the constant values termed as αα and ββ.
Figure 2. Graph depictsαα andββ using exponential approximation of Pr and Vr data.
Exponential approximation is employed to analyze the increasing nature of the pressure
value to the corresponding increase in the velocity value. From the graph, it is noted that α
constant is the slope of the exponential equation, and the y-intercept is calculated when x
is zero and the value is taken as β constant. The α constant 0.1857 kPa and the β constant
2.3 kPa values are observed from the curve and set as a standard constant for Equation (1)
α
Biosensors 2023, 13, 757
β
α
β
6 of 13
to calculate the blood pressure value computed from the pulse wave velocity embedded in
the non-invasive device CaNVAS.
Furthermore, the arterial diameter, arterial compliance,
ff
and stiffness index are computed with an appropriate equation, which is embedded in the
processor to find the vascular parameters
of the user using CaNVAS. Hereafter, the terms
tt
“user” and “attendant” refer to the one whose vascular or clinical parameters are specifically
measured using the non-invasive device. A sequential algorithm is carried out in the processor,
as shown in Figure 3, for the computation of vascular parameters.
Figure 3. Flowchart representing the process of CaNVAS.
Initially, user inputs are received from the choice of either data acquisition or validation.
The pre-processed received echoes are fed into the microprocessor and digitized, and the
reflected frequency (Fo ) is calculated by a frequency counter as step 1.
F F= −
F F− F(Hz)
Fs =
t
o
(3)
tz
In step 2, the Doppler shift Fs in kilo hertz (KHz) is calculated using Equation (3),
where Fs will be positive because the blood flows towards the receiver. The pulse wave
velocity, which comprises the peak systolic velocity (Vs ) and the end diastolic velocity (Vd ),
is computed in step 3 with the same procedure as carried out for finding the reference
velocity values using Equation (2). The blood pressure value is derived by applying the
velocity values obtained in step 3 and with the derived constants α (0.1857) and β (2.3). The
Biosensors 2023, 13, 757
7 of 13
systolic (Ps ) and diastolic blood pressures (Pd ) in kPa are calculated with the corresponding
systolic and diastolic velocity values by the two Equations (4) and (5).
Ps = αV2s + β (kPa)
(4)
Pd = αV2d + β (kPa)
(5)
The processor is programmed with a diameter equation to calculate the systolic arterial
diameter Ds and the diastolic arterial diameter Dd in millimeters (mm) of the target artery
of the user, using Equations (6) and (7).
Ds =
e(kPs ) t Eo
(mm)
ρv2s
(6)
Dd =
e(kPd ) t Eo
(mm)
ρv2d
(7)
The algorithm used in CaNVAS incorporates various constants. The arterial constant
(k) is set to 0.017. The thickness of the artery is predetermined based on the target artery:
0.25mm for the radial artery, 0.29 mm for the brachial artery, and 1.5mm for the carotid
artery. The density of blood (ρ) is assigned a value of 1060 kg/m3 . The variables Ps , Pd ,
Vs , and Vd represent the instantaneous systolic and diastolic blood pressures and the
velocity values obtained in previous steps. The initial elasticity of the artery (Eo ) is set
to 13.33 kPa, which will be adjusted according to each subject in subsequent steps. The
elasticity property of the artery is determined by its capacity to compress and extend
during systole and diastole in one cycle, as reflected in the frequency shift obtained from
the receiver. The ratio of the systolic and diastolic diameters is compared to the ratio
of the maximum frequency shift to the minimum frequency shift recorded during one
cycle. Any differences in these values are considered errors. If the error exceeds a certain
threshold, the elasticity value (Eo ) is modified from the initial set value, and step 5 is
repeated until the error is minimized. Consequently, the final value of Eo is used as the
appropriate elasticity value for the user. To ensure system performance, the non-invasive
device includes a checkpoint to verify the derived diameter using the diameter equation
against the respective arterial site. This derived diameter is correlated with the thickness–
diameter function (T/D), allowing for the detection of deviations beyond 0.3 ± 40%. If
the deviation of the thickness–diameter value exceeds 40%, it indicates improper device
positioning, and this information is displayed on the display module.
If the input received from the user matches the condition described in step 2, the device
will execute all the preceding steps exactly as it did for choice 1, including the validation
process. Otherwise, it will skip step 8. During the validation phase, only the blood pressure
parameter is validated using an external device. This parameter is crucial for calculating
the arterial diameter, arterial compliance, and stiffness index, which are the key parameters
used to validate the integrated device. The reference blood pressure value obtained from a
gold standard BP apparatus, as measured by the user, is compared with the values provided
by CaNVAS. The device will display the error value if it exceeds ±0.67 kPa, as is clinically
acceptable. Equations (8) and (9) are employed to derive the arterial compliance (AC) in
mm2 kPa− 1 and the stiffness index (SI), where ∆P is the difference between the systolic
and diastolic blood pressure parameters and ∆D is the change in the systolic and diastolic
diameters, calculated in the previous steps using CaNVAS.
π D2s − D2d
AC =
(mm2 kPa−1 )
(8)
4∆P
Biosensors 2023, 13, 757
8 of 13
ln PPs
d
SI =
∆D
Dd
(9)
The individual will utilize the user-friendly and non-invasive device CaNVAS to
continuously obtain the specified vascular parameter details without any time gaps. A
display unit will show the user’s ongoing measurements, including the instantaneous
systolic pressure (SYS), instantaneous diastolic pressure (DIA), peak systolic velocity (PSV),
end diastolic velocity (EDV), arterial compliance (AC), and stiffness index (SI). Additionally,
the device’s validation status will be displayed. The statistical analysis conducted in this
study utilized a linear regression model in Microsoft Excel 2016.
3. Results
For the performance analysis of CaNVAS, a group of 300 participants was selected.
This group consisted of both healthy individuals and individuals with a history of hypertension, hypotension, and cardiovascular disorders. Additionally, subjects undergoing regular
renal dialysis were included. The age range of the participants was 21 to 55 years, with an
average age of 38 ± 17. Among the participants, 65% were female and 35% were male. This
study conducted on an educational campus included a significant number of adolescent
students and faculty members as participants. To measure the vascular parameters and
blood pressure of these individuals, both CaNVAS and Diamond Deluxe BPMR-120 (Otica
Meditronix Co, Vadodara, India.) were used simultaneously. The institute’s ethics committee approved this study, and all participants were provided with a clear explanation
of the experimental procedure. Informed consent was obtained from each participant
before data acquisition. The typical time required for one set of readings was under two
minutes. All subjects underwent comprehensive measurements. In certain cases, a few
additional minutes were necessary for the initial device set-up to accurately trace the artery
due to the associated challenge. The BPMR-120 cuff was applied to the brachial artery,
ensuring it covered at least 80% of the arm’s circumference. CaNVAS was positioned over
the target artery (radial artery) using the self-evaluation position correcting unit. During
the assessment, all subjects were seated in a relaxed position with their feet resting on the
floor and their arms supported at heart level. The measurements were repeated for the
same subjects after engaging in physical exercise (Rapid Steps Climbing). Two trials were
conducted under each condition to assess the repeatability of CaNVAS. Blood pressure
data obtained from CaNVAS were compared to those measured using the gold standard
blood pressure device, Diamond Deluxe BPMR-120. Among the total of 600 data points
collected for both conditions (normal and after exercise), a few data points were considered
outliers due to movement artifacts and powerline interference. These suspected cases of
erratic data differences were excluded from the analysis, resulting in a remaining dataset of
556 data points, which accounted for 90% of the total.
3.1. Linear Regression Analysis
Figure 4 provides an overview of the linear regression analysis conducted on the pressure
values obtained from two measurement devices under normal and post-exercise conditions.
To ensure compatibility between the devices, a conversion factor (7.501) was applied to
convert mmHg to kPa units. The blood pressure results from CaNVAS demonstrated a strong
positive correlation with the corresponding values obtained from BPMR-120, which were
recorded as whole numbers. Correlation coefficients (r) of 0.94 and 0.92 were achieved for
the pressure values measured under normal and post-exercise conditions, respectively. The
slightly lower correlation observed after exercise can be attributed to the short time delay
involved in setting up the device on the target artery, resulting in sequential measurements
rather than simultaneous ones during data recording after exercise.
Biosensors 2023, 13, 757
9 of 13
(a)
(b)
Figure 4. Comparison of blood pressure (BP) measurements from CaNVAS with those obtained from
BPMR-120: (a) during resting condition; (b) after performing physical exercise.
In order to evaluate the agreement between blood pressure values obtained from
CaNVAS and the gold standard device under two conditions, Bland–Altman plots were
generated with limits of agreement set at ±2 standard deviations (SD), as depicted in
Figure 5. The plots reveal a substantial level of agreement between the two measurement
devices, with a bias value close to zero. This indicates the reliability of the method employed
in CaNVAS for estimating blood pressure based on directly measured pulse wave velocity.
Figure 5. Bland–Altman plot of BP values measured from CaNVAS and BPMR-120 during both
resting and after performing physical exercise.
ff
The mean difference
between the blood pressure values obtained from the two devices
ff
was observed to be −0.123−± 2.756, indicating minimal distortion in measurements. The
−
ff in measurements are randomly
graphical representation illustrates that the differences
ff
distributed, with 98% of the data falling within the range of agreement.
The correlation matrix is estimated within the vascular parameters (for instance, the
Mean Arterial Pressure (MAP), Mean Pulse Wave Velocity (MPWV), Mean Diameter (MD),
arterial compliance (AC), and stiffness index
ff (SI)) calculated using CaNVAS for all subjects.
ff
The correlation coefficients value
ffi depicted in Figure 6 shows a strong relationship between
ffi
the parameters of the individuals. A positive correlation is shown for the pairs (i) MAP
Biosensors 2023, 13, 757
10 of 13
with MPWV, MD, and AC; (ii) MPWV with SI, AC, and MD; and (iii) MD with SI and AC,
which implies that a slight increase in one value will be reflected in another, which acts as
the biomarker for diagnosing vascular abnormalities. There is a strong negative correlation
between the subject’s SI and MAP and AC, which indicates an increase in the Mean Arterial
blood pressure value, and arterial compliance decreases the vascular stiffness; this indicates
that the
ff forecast estimation methodology carried out in CaNVAS to measure the PWV, BP,
AC, and SI is trustworthy.
ffi coefficients calculated between vascular
Figure 6. Correlation matrix representing the correlation
parameters within the subjects.
3.2. Repeatability Analysis
During this study, two trials were conducted to assess the repeatability of CaNVAS.
The frequencyffof differences between the two trials for blood pressure (BP), pulse wave
velocity (PWV), arterial compliance (AC), ffand stiffness index (SI) values is depicted in
ffi coefficient of variability (repeatability), the standard deviation
Figure 7. To determine the
ff
of the differences between the two CaNVAS measurements is divided by the average of
ffi
the mean. The coefficient of repeatability for the parameters is presented in Table 1. This
demonstrates the device’s capability to provide accurate evaluations of vascular parameters.
ffi
ffi
−
−
ff
Figure 7. Graph showing the frequency of the error obtained during the measurement of BP, MPWV,
AC, and SI in two trials.
Biosensors 2023, 13, 757
11 of 13
Table 1. Coefficient of repeatability for CaNVAS measurements.
Parameters Measured
Coefficient of Repeatability in %
Blood Pressure (kPa)
Pulse Wave Velocity (ms−1 )
Arterial Compliance
(mm2 kPa−1 )
Stiffness Index
10.2
11.3
20.9
7.6
4. Discussion
This paper presents a novel device called CaNVAS, along with its methodology, for
non-invasively evaluating continuous vascular parameters. CaNVAS utilizes ultrasound
technology to measure pulse wave velocity (PWV) based on the reflected frequency shift
signal from the artery. This method shows promise and is employed for deriving blood
pressure (BP) as well as for estimating arterial compliance (AC) and stiffness index (SI).
In a previous device developed by Drzewiecki et al. (1992), an arterial tonometer was
used to non-invasively measure continuous blood pressure by identifying the center of
the blood vessel [39]. The device required the deflection area of the artery vessel to make
contact with the sensor surface for optimal deflection, necessitating a specially designed
positioning unit. In contrast, CaNVAS, the proposed device in this study, records ultrasound
frequency shifts when placed in proximity to the blood vessel, thereby overcoming the need
to involve the entire blood vessel area and the interference angle for observing relevant
changes. Another device developed by Jeyaraj et al. (2015) measured arterial compliance
and stiffness parameters non-invasively using ultrasound principles [40]. However, this
device relied on a one-time measurement of blood pressure using an external device. In
comparison, CaNVAS has the advantage of simultaneously measuring instantaneous blood
pressure, arterial compliance, and stiffness index, making it a superior solution. Overall,
the proposed CaNVAS device offers a novel approach for the non-invasive measurement of
continuous vascular parameters, including blood pressure, arterial compliance, and stiffness
index, utilizing ultrasound technology and overcoming the limitations of existing devices.
This paper provides a comprehensive description of the hardware and algorithm of
the CaNVAS device. The performance of the device was assessed using approximately
300 subjects from various age groups, ranging between 18 and 35 years old, with an average
age of 26 ± 9. The blood pressure values measured by CaNVAS were compared to those
obtained from a sphygmomanometer (BPMR-120), demonstrating an accuracy rate of 95%.
The reliability of other parameters, such as arterial compliance (AC) and stiffness index
(SI), measured by CaNVAS was evaluated using a self-correlation technique. Because these
parameters are interconnected with blood pressure, even a slight error in blood pressure
measurement can impact AC and SI values. It was observed that AC and SI measured by
CaNVAS are influenced by age. Similar to the findings of Jayaraj et al. in their ARTSENS
device, there is a gradual decrease in arterial compliance and an increase in stiffness index
with advancing age.
To ensure the repeatability of CaNVAS, linear regression analysis and Bland–Altman plots
were employed. The results demonstrated an average coefficient of variability of 12.5%, which
is consistent with previous research and indicates a comparable level of repeatability.
5. Conclusions
The obtained results provide strong evidence that CaNVAS is a promising device
for measuring vascular parameters comparable to existing modalities. Its compact size,
affordability, repeatability, and user-friendly nature make CaNVAS suitable for medical
applications, such as measuring continuous blood pressure during dialysis procedures and
monitoring the health status of subjects after cardiac surgery. However, it is important to
note the limitations of this study. The hysteresis behavior of psychobiological variables
during exercise was not considered, and this aspect will be investigated in future research.
Additionally, the study population mainly consisted of adolescent students and middle-
Biosensors 2023, 13, 757
12 of 13
aged faculty members due to its location on an educational campus. This demographic
bias represents a limitation, and a demographic analysis of CaNVAS will be explored in
future studies.
Author Contributions: Conceptualization, M.Y.S., B.M. and S.P.; methodology, I.A., A.A. and S.A.A.;
investigation, S.B., A.A., S.A.A. and A.A.A.; writing—original draft preparation, B.M., A.A.A. and
M.Y.S.; writing—review and editing, S.B. and A.A.; project administration, I.A., A.A. and S.A.A.; funding
acquisition, M.Y.S. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: This study was conducted in accordance with the Rajalakshmi Engineering College and approved by the Institutional Review Board (or Ethics Committee) of
Rajalakshmi Engineering College (approval no: REC/RAB/2019).
Informed Consent Statement: Informed consent was obtained from all subjects involved in this study.
Data Availability Statement: The clinical data are available upon request in accordance with the
volunteers’ informed consent. The data will not be shared online.
Acknowledgments: The authors extend their appreciation to the Deputyship of Research and Innovation, Ministry of Education in Saudi Arabia for funding this research work through the Project
Number IFP-2022-05.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
Dart, A.M.; Lacombe, F.; Yeoh, J.K.; Cameron, J.D.; Jennings, G.L.; Laufer, E.; Esmore, D.S. Aortic distensibility in patients with
isolated hypercholesterolaemia, coronary artery disease, or cardiac transplant. Lancet 1991, 338, 270–273. [CrossRef] [PubMed]
Hirai, T.; Sasayama, S.; Kawasaki, T.; Yagi, S. Stiffness of systemic arteries in patients with myocardial infarction. A noninvasive
method to predict severity of coronary atherosclerosis. Circulation 1989, 80, 78–86, Erratum in: Circulation 1989, 80, 1946.
[CrossRef] [PubMed]
Zoungas, S.; Asmar, R.P. Arterial stiffness and cardiovascular outcome. Clin. Exp. Pharmacol. Physiol. 2007, 34, 647–651. [CrossRef]
[PubMed]
Blacher, J.; Pannier, B.; Guerin, A.P.; Marchais, S.J.; Safar, M.E.; London, G.M. Carotid arterial stiffness as a predictor of
cardiovascular and all-cause mortality in end-stage renal disease. Hypertension 1998, 32, 570–574. [CrossRef] [PubMed]
Gatzka, C.D.; Cameron, J.D.; Kingwell, B.A.; Dart, A.M. Relation between coronary artery disease, aortic stiffness, and left
ventricular structure in a population sample. Hypertension 1998, 32, 575–578. [CrossRef]
Lehmann, E.D.; Hopkins, K.D.; Jones, R.L.; Rudd, A.G.; Gosling, R.G. Aortic distensibility in patients with cerebrovascular
disease. Clin. Sci. 1995, 89, 247–253. [CrossRef]
Pannier, B.M.; Avolio, A.P.; Hoeks, A.; Mancia, G.; Takazawa, K. Methods and devices for measuring arterial compliance in
humans. Am. J. Hypertens. 2002, 15, 743–753. [CrossRef]
Komine, H.; Asai, Y.; Yokoi, T.; Yoshizawa, M. Non-invasive assessment of arterial stiffness using oscillometric blood pressure
measurement. Biomed. Eng. Online 2012, 11, 6. [CrossRef]
Whitaker, T.; Wilson, S.C.; Spang, K.; Wawro, T.J.; Lane, J.A. Motion Management in a Fast Blood Pressure Measurement Device.
U.S. Patent 7,429,245, 30 September 2008.
Chen, Y.; Li, L.; Hershler, C.; Dill, R.P. Continuous Non-Invasive Blood Pressure Monitoring Method and Apparatus. U.S. Patent
6,599,251, 29 July 2003.
Frankenreiter, M.; Rometsch, R.; Seher, J.P. HP Inc, Blood Pressure Monitor. U.S. Patent 4,953,557, 4 September 1990.
Forstner, K.; Yen, G.F.A.C.Y. Non Invasive Blood Pressure Monitor and a Method for the Non-Invasive Measurement of the Blood
Pressure. U.S. Patent 6,447,457, 10 September 2002.
Vatner, S.F.; Zhang, J.; Vyzas, C.; Mishra, K.; Graham, R.M.; Vatner, D.E. Vascular Stiffness in Aging and Disease. Front. Physiol.
2021, 12, 762437. [CrossRef]
Hoctor, R.T.; Thomenius, K.E.; Dentinger, A.M.; Mccarter, J.W. Method and Apparatus for Ultrasonic Continuous, Non-Invasive
Blood Pressure Monitoring. U.S. Patent 7,125,383, 24 October 2006.
Kuri, Y. System and Method for Determining Arterial Compliance and Stiffness. U.S. Patent 9,408,541, 9 August 2016.
Xu, M. Local Measurement of the Pulse Wave Velocity Using Doppler Ultrasound. Master’s Thesis, Massachusetts Institute of
Technology, Cambridge, MA, USA, 24 May 2002.
Polańczyk, A.; Woźniak, T.; Strzelecki, M.; Szubert, W.; Stefańczyk, L. Evaluating an algorithm for 3D reconstruction of blood vessels
for further simulations of hemodynamic in human artery branches. In Proceedings of the 2016 Signal Processing: Algorithms,
Architectures, Arrangements, and Applications (SPA), Poznan, Poland, 21–23 September 2016; pp. 103–107. [CrossRef]
Biosensors 2023, 13, 757
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
13 of 13
Polańczyk, A.; Strzelecki, M.; Woźniak, T.; Szubert, W.; Stefańczyk, L. 3D Blood Vessels Reconstruction Based on Segmented
CT Data for Further Simulations of Hemodynamic in Human Artery Branches. Found. Comput. Decis. Sci. 2017, 42, 359–371.
[CrossRef]
Polanczyk, A.; Podgorski, M.; Polanczyk, M.; Veshkina, N.; Zbicinski, I.; Stefanczyk, L.; Neumayer, C. A novel method for
describing biomechanical properties of the aortic wall based on the three-dimensional fluid-structure interaction model. Interact.
Cardiovasc. Thorac. Surg. 2019, 28, 306–315. [CrossRef]
Riley, W.A., Jr.; Barnes, R.W.; Schey, H.M. An approach to the noninvasive periodic assessment of arterial elasticity in the young.
Prev. Med. 1984, 13, 169–184. [CrossRef]
Ahlgren, A.R.; Hansen, F.; Sonesson, B.; Länne, T. Stiffness and diameter of the common carotid artery and abdominal aorta in
women. Ultrasound. Med. Biol. 1997, 23, 983–988. [CrossRef]
Gamble, G.; Zorn, J.; Sanders, G.; MacMahon, S.; Sharpe, N. Estimation of arterial stiffness, compliance, and distensibility from
M-mode ultrasound measurements of the common carotid artery. Stroke 1994, 25, 11–16. [CrossRef]
Kawasaki, T.; Sasayama, S.; Yagi, S.; Asakawa, T.; Hirai, T. Non-invasive assessment of the age related changes in stiffness of
major branches of the human arteries. Cardiovasc. Res. 1987, 21, 678–687. [CrossRef]
Magda, S.L.; Ciobanu, A.O.; Florescu, M.; Vinereanu, D. Comparative reproducibility of the noninvasive ultrasound methods for
the assessment of vascular function. Heart Vessels 2013, 28, 143–150. [CrossRef]
Nichols, W.W.; O’Rourke, M.F.; Kenney, W.L. McDonald’s Blood Flow in Arteries: Theoretical, Experimental and Clinical Principles, 7th
ed.; CRC Press: Boca Raton, FL, USA, 2022.
Nagai, Y.; Fleg, J.L.; Kemper, M.K.; Rywik, T.M.; Earley, C.J.; Metter, E.J. Carotid arterial stiffness as a surrogate for aortic stiffness:
Relationship between carotid artery pressure-strain elastic modulus and aortic pulse wave velocity. Ultrasound. Med. Biol. 1999,
25, 181–188. [CrossRef]
Peterson, L.H.; Jensen, R.E.; Parnell, J. Mechanical Properties of Arteries in Vivo. Circ. Res. 1960, 8, 622–639. [CrossRef]
Xing, X.; Ma, Z.; Zhang, M.; Zhou, Y.; Dong, W.; Song, M. An Unobtrusive and Calibration-free Blood Pressure Estimation Method
using Photoplethysmography and Biometrics. Sci. Rep. 2019, 9, 8611. [CrossRef]
Zheng, D.; Yao, Y.; Morrison, I.; Greenwald, S. Photoplethysmographic assessment of arterial stiffness and endothelial function.
In Photoplethysmography; Academic Press: London, UK, 2022; pp. 235–276.
Danpinid, A.; Luo, J.; Vappou, J.; Terdtoon, P.; Konofagou, E.E. In vivo characterization of the aortic wall stress-strain relationship.
Ultrasonics 2010, 50, 654–665. [CrossRef]
Arnett, D.K.; Evans, G.W.; Riley, W.A. Arterial stiffness: A new cardiovascular risk factor? Am. J. Epidemiol. 1994, 140, 669–682.
[CrossRef] [PubMed]
Arnett, D.K.; Chambless, L.E.; Kim, H.; Evans, G.W.; Riley, W. Variability in ultrasonic measurements of arterial stiffness in the
Atherosclerosis Risk in Communities study. Ultrasound. Med. Biol. 1999, 25, 175–180. [CrossRef] [PubMed]
Bussy, C.; Boutouyrie, P.; Lacolley, P.; Challande, P.; Laurent, S. Intrinsic stiffness of the carotid arterial wall material in essential
hypertensives. Hypertension 2000, 35, 1049–1054. [CrossRef] [PubMed]
Claridge, M.W.; Bate, G.R.; Hoskins, P.R.; Adam, D.J.; Bradbury, A.W.; Wilmink, A.B. Measurement of arterial stiffness in subjects
with vascular disease: Are vessel wall changes more sensitive than increase in intima-media thickness? Atherosclerosis 2009, 205,
477–480. [CrossRef] [PubMed]
Nogata, F.; Yokota, Y.; Kawamura, Y.; Walsh, W.R. System for estimating sclerosis of in vivo arteries based on ultrasound B-mode
image analysis. In Proceedings of the 10th International Conference on Control, Automation, Robotics and Vision, AIP Conference
Proceedings, Hanoi, Vietnam, 17–20 December 2008; pp. 1779–1782.
Selzer, R.H.; Mack, W.J.; Lee, P.L.; Kwong-Fu, H.; Hodis, H.N. Improved common carotid elasticity and intima-media thickness
measurements from computer analysis of sequential ultrasound frames. Atherosclerosis 2001, 154, 185–193. [CrossRef]
Mackenzie, I.S.; Wilkinson, I.B.; Cockcroft, J.R. Assessment of arterial stiffness in clinical practice. Qjm 2002, 95, 67–74. [CrossRef]
Luo, J.; Konofagou, E. A fast normalized cross-correlation calculation method for motion estimation. IEEE Trans. Ultrason.
Ferroelectr. Freq. Control 2010, 57, 1347–1357. [CrossRef]
Drzewiecki, G.M.; Butterfield, R.D.; Ciaccio, E.J. Rutgers State University of New Jersey, Pressure Waveform Monitor. U.S. Patent
5,363,855, 14 November 1994.
Joseph, J.; Radhakrishnan, R.; Kusmakar, S.; Thrivikraman, A.S.; Sivaprakasam, M. Technical Validation of ARTSENS-An Image
Free Device for Evaluation of Vascular Stiffness. IEEE J. Transl. Eng. Health Med. 2015, 3, 1900213. [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.