Indices and Repeatability Tests of Cardiovascular Function Performed

on the Arterial Distension Waveform

Case Study: Angiography Intervention

V. G. Almeida

, J. Borba

, H. C. Pereira

1,2

, T. Pereira

, J. Cardoso

and C. Correia

Instrumentation Center (GEI-CI), University of Coimbra, Coimbra, Portugal

ISA-Intelligent Sensing Anywhere, Coimbra, Portugal

Keywords:

Arterial Distension Waveform, Augmentation Index, Clinical Validation, Repeatability Trials, Angiography

Intervention.

Abstract:

The arterial distension waveform (ADW) analysis is a reliable technique for cardiovascular function assess-

ment. The purpose of this study was to perform the pre-clinical validation of a non-invasive prototype focusing

the repeatability tests and cross-relationships between different subject groups. The evaluation focused param-

eters retrieved from ADW: systolic peak (SP), dicrotic notch (DN), RP (reﬂection point) and Augmentation

index (AI). One hundred and ﬁfty one subjects (61 men and 90 women, aged between 18 and 80 years) were

assigned into four groups based on their clinical characteristics. Database is constituted by healthy, hyperten-

sive and subjects that suffer from stenosis. The cross-correlations analysis between groups allows establishing

time parameterizations for each one. Furthermore, the differences between the left and right carotid artery

suggest intrinsically variability for each one of the subjects. The coefﬁcient of variation (CV) mean value

obtained for all measurements was 18.58%, maximum rate of 33.7% and minimum 8.9%. The stenosis case

study demonstrate the potentialities of the use of this prototype in the detection of cardiac anomalies by the

monitoring of state alterations through RP, SP and DN time parameterizations with visible changes in RP and

SP values (after carotid intervention RP appears later than SP, in opposition with values before intervention),

while DN associated time changes little. The tests performed on the ADW showed that is possible the reliable

measurement of morphological patterns changes.

1 INTRODUCTION

The prevention and treatment guidelines of cardio-

vascular disease (CVD) have focused on the modi-

ﬁcation of risk factors, such as hypertension, smok-

ing or hyperglycaemia, among others that could po-

tentially reduce cardiac events. Furthermore, the

early non-invasive identiﬁcation of hemodynamic al-

terations that result from pathological situations can

facilitate much more precise risk stratiﬁcation.

Arterial walls are the primary site of disease in

arteriosclerotic vascular disease, and their properties

and function are commonly affected by the risk fac-

tors above referred. So, the assessment of wall in-

tegrity can provide an accurate prediction of car-

diovascular (CV) risk (Hamilton et al., 2007). CV

risk assessment can be performed using several risk

tools based upon multi-variable equations. In spite

of the difﬁculty in absolute risk prediction, they are

extremely useful in assessing or estimating risk, as

well as in prioritising treatments. The most used are

the Framingham (Bitton and Gaziano, 2010) and the

SCORE (Conroy, 2003).

Some factors have been focused by scientiﬁc com-

munity due to their relationship with risk occurrences.

The INTERHEART study established the smoking

status, history of hypertension or diabetes, waist hip

ratio, dietary pattern, physical activity, alcohol con-

sumption, blood apolipoproteins and psychosocial

factors collectively accounting for more than 90%

of the risk in the acute myocardial infarction (Yusuf

et al., 2004). Serum total, high density (HDL) choles-

terol and serum triglyceride are used in the most of

risk scores. However, HDL is frequently ignored by

some authors that emphasize its poor sensitivity for

the identiﬁcation of individuals at high risk (Jones

et al., 2001). Hypertension is also reported as one of

the most important factors, and should not be ignored.

Arterial stiffness is one of the early detectable symp-

toms of adverse structural and functional changes

138

G. Almeida V., Borba J., C. Pereira H., Pereira T., Cardoso J. and Correia C..

Indices and Repeatability Tests of Cardiovascular Function Performed on the Arterial Distension Waveform - Case Study: Angiography Intervention.

DOI: 10.5220/0004219301380144

In Proceedings of the International Conference on Bio-inspired Systems and Signal Processing (BIOSIGNALS-2013), pages 138-144

ISBN: 978-989-8565-36-5

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

Figure 1: Database groups and signal processing tasks per-

formed for each one.

within the vessel wall (Laurent, 2006). A wide vari-

ety of non-invasive techniques are employed to assess

its development (Cavalcante et al., 2011; Mackenzie

et al., 2002).

The arterial distension waveform (ADW) is an in-

teresting signal that can be used in the analysis of rele-

vant morphological features (Avolio et al., 2010). The

major determinants are systolic peak (SP), reﬂected

wave (RP) and dicrotic notch (DN) that results from

the aortic valve closure at end of systole. The pres-

ence of branches and tapers in the arterial tree leads

to occurrence of reﬂection waves that are inﬂuenced

by the arterial wall stiffening. The increase of arterial

stiffness leads to early occurrence of RP rather than

SP leading to the augmentation phenomenon. The

augmentation index (AI) is used to quantify this fact,

and is calculated as a percentage of the increment of

pressure to the pulse pressure (Hamilton et al., 2007).

Little is known about the cross-relationships of

these indices in different risk groups. The available

studies compare within the same subject (Huck et al.,

2007) several techniques such as, tonometry, ultra-

sound, Doppler and brachial ﬂow-mediated dilation

(Liang et al., 1998). Furthermore, the analysis of

ADW from different anatomical origins, such as left

and right carotid artery could be interesting in the

analysis of anatomical differences (Luo et al., 2011).

The non-invasive prototype to be tested was devel-

oped in a previous work for distension wave estima-

tion based on piezoelectric (PZ) sensors. A set of pre-

liminary tests were done in ﬂexible tubes and in vol-

unteers to physiological monitoring (Almeida et al.,

2011a; Almeida et al., 2011b). The main focus of this

work is performing a set of clinical tests, as follow:

• Exploration of the cross-correlations between

subjects;

• Perform a set of repeatability tests in successive

measurements and distinguish the differences be-

tween left and right carotid ADW;

• Evaluate the capability of probe in detect mod-

iﬁcations in clinical states by the monitoring of

ADW after and before angiography procedure, us-

ing an invasive method as reference.

The selection and description of primary variables

and groups are issued in section 2. Results are pre-

sented in section 3. Finally, discussion and conclu-

sions are stated in sections 4 and 5, respectively.

2 METHODS

This section describes the dataset and protocol used

during data acquisition. The database details are de-

picted in Figure 1 (below row).

2.1 Subjects

One hundred and ﬁfty one volunteers, 61 male and

90 female, aged between 18 and 80 years were in-

cluded in this study. Subjects were recruited from

Coimbra University Hospital Centre (C.H.U.C.) and

advertisements placed in public platforms calling for

healthy volunteers. Informal consent was obtained af-

ter full explanation of the purpose, nature and risk of

all procedures used. The research was approved by

the Committees of the Coimbra University Hospital

Centre. The database is composed by four indepen-

dent groups, whose demographic data are listed in Ta-

ble 1.

• Group I is composed by 114 normal subjects (be-

tween 18 and 30 years).

• Group II is composed by 12 normal subjects be-

tween 21 and 29 years.

• Group III is composed by 6 subjects that suffer

from severe stenosis.

• Group IV comprises 19 hypertensive subjects.

IndicesandRepeatabilityTestsofCardiovascularFunctionPerformedontheArterialDistensionWaveform-CaseStudy:

AngiographyIntervention

139

Table 1: Demographic data comparison for each one of groups analysed.

Variable Group I Group II Group III Group IV

Age 21.90 ± 3.32 23.50 ± 2.43 72.5 ± 5.44 58.63 ± 12.51

Sex (M/F) 45/69 4/8 4/2 8/11

Smoker (Y/N) 10/104 0/12 M.D.* 2/17

Diabetes (Y/N) 0/114 0/12 M.D.* 2/17

Weight (Kg) 62.46 ± 10.24 59.66 ± 10.96 M.D.* 74.42 ± 10.77

Height(m) 1.69±0.09 1.66±0.06 M.D.* 1.64 ± 0.10

BMI(Kg/m) 21.80 ± 2.59 21.45 ± 2.71 M.D.* 27.81 ± 5.28

SBP(mmHg) 108.78 ± 11.59 105.25 ± 5.86 144.33 ± 40.08 155.42 ± 27.32

DBP(mmHg) 69.52 ± 7.84 65.67 ± 6.71 81 ± 18.34 90.95 ± 17.45

HR(beats/min) 70.11 ± 10.83 71.17 ± 10.64 72.5 ± 3.53 63.79 ± 6.08

Note: Data were expressed in mean±standard deviation (SD)

*M. D. -Missing data

2.2 Protocol

ADWs were recorded at the sampling rate of 1kHz

using a non-invasive PZ probe previously developed

(Almeida et al., 2011b). The probe is placed over

the carotid artery, and is held by a collar to avoid

noise interference and artefacts that may arise from

the interaction between the probe and the operators

hand. Age, sex, weight, height, smoking habits and

diabetes history were registered for each subject. The

Body Mass Index (BMI) was later calculated. Sys-

tolic blood pressure (SBP), diastolic blood pressure

(DBP) and heart rate (HR) values were measured in

the left arm with an automated digital oscillomet-

ric sphygmomanometer (Omron M6 Comfort, Kyoto,

Japan). The following requirements were performed

for each one of groups:

• Group I- One measurement (three trials) was

taken for each subject during 30-40 seconds. The

subjects remained quiet and seated on a comfort-

able chair;

• Group II- For each subjects two successive ses-

sions were monitored. In each session, 3 trials

of 30-40 seconds were performed for each carotid

site (right and left). All measurements were made

in a similar time of the day and at the same tem-

perature controlled room (22-23

◦

Celsius), to min-

imize the climatic variation. The subjects re-

mained quiet and seated on a comfortable chair;

• Group III- These subjects were monitored under

carotid intervention (due to a stenosis that partly

blocks the artery blood ﬂow). The angiography is

the imaging method of choice to visualize it. After

local anesthesia, a surgical cut is executed next to

the groin, and after a ﬁlter device is opened above

the lesion, the stent is implanted. The balloon is

then inﬂated at the stenosis plaque level, decreas-

ing the stenosis after deﬂation, and the ﬁlter de-

vice is kept open as prevention for an eventual em-

bolic trapping. With this surgical technique, the

blood ﬂow was restored to the normal values since

the diameter of the vessel enlarged to the same im-

posed by the stent. The ADW was sequentially

collected by our probe and invasive equipment,

Axiom Sensis (Siemens). Only a small segment

of 3-4 seconds was analyzed.

• Group IV- The data from group IV were ac-

quired during hospitalization. One measurement

(three trials) was taken for each subject during 30-

40 seconds. The subjects were monitored prior

to taking any medication and subjects remained

quiet and seated on a comfortable chair.

2.3 Signal Acquisition and Discrete

Time Signal Processing

Data processing was undertaken in MATLAB 2011b

and statistical analysis was performed using Mi-

crosoft Excel 2010 and SPSS18.0 statistical soft-

ware. The collected data consisted in demographic

and pulse waveform parameters. The overall list of

demographic parameters (expressed as mean±S.D.)

is described in Table 1.

The methodology adopted in the pulse wave anal-

ysis consists in the removal of bad pulses that may re-

sults from voluntary or involuntary movements, such

as respiration. The criteria used to remove these

pulses consisted in the analysis of amplitude and

width abnormal pulse variations. Then, the pulse-

by-pulse analysis was performed, consisting in the

identiﬁcation of the systolic peak (SP), diastolic notch

(DN) and reﬂection point (RP) in the ADW, such rep-

resented in Figure 1 (upper row). An algorithm for

prominent points identiﬁcation was previously devel-

oped (Almeida et al., 2011a) for this task. This algo-

rithm action is based on the analysis a number of fea-

tures of the arterial distension waveform and its ﬁrst

BIOSIGNALS2013-InternationalConferenceonBio-inspiredSystemsandSignalProcessing

140

derivative. After acquisition, the pulses are subject to

a segmentation process and amplitude normalization

to the diastolic-systolic pressure interval in a manner

to allow a reliable comparison among subjects.

2.4 Data Analysis

2.4.1 Cross-relationships

SP, DN and RP associated time and amplitude was

computed for all of the groups, as well as AI values.

Group I and II are both composed by healthy subjects.

To avoid similar patterns, only Group I was used in

the analysis of cross-relationships. The arrival time

histogram distributions were computed for the Groups

I, III and IV, shown in Figure 2.

2.4.2 Pulse Variability

Group II was used to study of the data repeatability.

The SP, DN and RP associated time and amplitude

information was used for this purpose by the analysis

of successive sessions and comparison of mean values

and standard deviations. Furthermore the coefﬁcient

of variation (CV) within each subject in the same ses-

sion was also computed. The left and right carotid

differences were analysed. The time and amplitude

parameterizations were also studied between left and

right carotid side, as well as the AI values.

2.4.3 Physiological Changes During

Angiography

The waveforms, before and after the procedure were

collected through the invasive and non-invasive de-

vices. However the comparison is only possible

through visual inspection due to the limitations in data

availability of invasive device.

3 RESULTS

3.1 Inter-groups Relationships

Figure 2 shows the arrival time histogram distribu-

tions, between all groups for the SP, DN and RP. SP

occurs later in hypertensive subjects (Group IV) con-

trasting with RP that occurs early in group IV and

later in group I. DN arrival time is similar in groups

I and IV (hypertensive and normal subjects). How-

ever, DN presents a peculiar behaviour characterized

by early time parameterizations in group III. When

the algorithm is not able to identify the prominent

Figure 2: SP, DN, RP cross-relationships between groups I,

II and IV.

points due to morphological artefacts or algorithm er-

rors, a null timing is assumed, that is visible in the

ﬁrst bin of each histogram.

3.2 Pulse Variability

The pulse variability between sessions was studied for

group II. The results are shown in Table 2, where

for each one of the parameters was computed the

mean ± SD.

Results demonstrate good repeatability obtained

from RP analysis, 168.90 ± 40.12 ms and in ampli-

tude 0.86±0.09 (a.u.) in session I and 166.41 ±34.10

ms and 0.87 ± 0.07 (a.u.) in session II. DN also

IndicesandRepeatabilityTestsofCardiovascularFunctionPerformedontheArterialDistensionWaveform-CaseStudy:

AngiographyIntervention

141

Table 2: Comparison of hemodynamic parameters obtained between sessions and bilateral common carotid acquisitions.

Parameter Session 1 Session 2 Left Right

SP (time(s)) 155.69 ± 58.81 177.74 ± 64.26 163.70 ± 60.51 169.27 ± 64.51

SP (amp(a.u)) 0.99 ± 0.01 0.99 ± 0.01 0.99 ± 0.01 0.99 ± 0.01

DN (time(s)) 285.98 ± 61.72 293.23 ± 39.31 286.96 ± 52.40 292.39 ± 51.98

DN (amp(a.u)) 0.70 ± 0.15 0.72 ± 0.13 0.72 ± ±0.15 0.71 ± 0.14

RP (time(s)) 168.90 ± 40.12 166.41 ± 34.10 169.18 ± 39.48 165.98 ± 34.66

RP (amp(a.u)) 0.86 ± 0.09 0.87 ± 0.07 0.88 ± 0.07 0.86 ± 0.08

AI (%) −2.67 ± 15.94 0.48 ± 13.84 −1.55 ± 14.26 −0.70 ± 15.91

Figure 3: A: AI distribution in function of carotid side of measurement for each one of the subjects.

presents good agreement between time and amplitude

information 285.98 ± 61.72 ms and 0.70±0.15 (a.u.)

in opposition with 293.23 ± 39.31 and 0.72 ± 0.13

(a.u.). SP presents more dispersion in time measure-

ments. Since SP amplitude was previously normal-

ized, during pre-processing, its relevance in the anal-

ysis is negligible.

The CV for each one of the subjects within the

same measurement, independent of the visit and

carotid side were computed. These values suggest in-

trinsically variability that needs to be taking in consid-

eration for each one of the subjects. The CV mean ob-

tained for all measurements was 18.58%, maximum

rate of 33.7% and minimum 8.9%. All of the CV val-

ues are displayed in Table 3. The left-right carotid

differences were also studied. The analysis of time

parameterizations (mean ± SD) is satisfactory, with-

out signiﬁcant differences, as is presented in Table 2.

Furthermore, for each one of the subjects were plotted

the AI values depending of carotid side of measure-

ment. It is evident AI dispersion for some of the sub-

jects. The less marked differences in AI distribution

are achieved for subjects 3 and 9, which are according

with lowest CV values. The largest differences in AI

and CV values occur for the 12 subject.

Table 3: CV values obtained for each subjects indepen-

dently of session and carotid side of measurement.

Subjects CV (%)

1 26.1

2 14.7

3 8.9

4 33.3

5 9.2

6 17.8

7 14.7

8 21.9

9 11.8

10 13.6

11 17.2

12 33.7

3.3 Angiography Subjects

The ADW comparison between both methods (inva-

sive and non-invasive) is shown in Figure 4. A small

segment (3-4 s) was chosen from each method before

and after carotid intervention. It is possible to note the

physiological alterations in both methods after carotid

intervention. The DN, RP, and SP associated time

BIOSIGNALS2013-InternationalConferenceonBio-inspiredSystemsandSignalProcessing

142

Figure 4: Set of three-four pulses detected invasively and non-invasively after and before carotid intervention. RP arrival time

for each one is represented by gray and blue circles, corresponding respectively to ADW collected before and after carotid

intervention.

Figure 5: SP, RP and DN time parameterization comparison

before and after carotid intervention.

were measured for the non-invasive method before

and after carotid procedure and the data are presented

in Figure 5. DN associated time changed little, as ex-

pected due to the absence of known cardiac valves

complications. However, visible changes occur in the

RP and SP analysis. Before the catheterization pro-

cedure, the RP occurs early, while SP appears later

but after the carotid intervention this tendency was in-

verted leading to the occurrence of normal ADW pat-

terns characteristics from healthy subjects (SP occurs

ﬁrst, and then RP). This fact seems to indicate that

the anomalous reﬂection waveform disappeared with

the carotid intervention, as indicated by blue circles in

Figure 4. The main purpose of this case study was to

prove that our probe is able to monitor physiological

alterations, such occurs in stenosis situations.

4 DISCUSSION

The analysis of pulse variability for ADW measure-

ments showed good results. There were no signicant

differences within each pair over the range of mea-

surements for any of the variables. The AI values

observed for each one of the subjects shows differ-

ences in mean value for the subjects (#2, #8, #12), but

correlated with high CV value leads to the assump-

tion that physiological variations are the main cause.

These results are according (Luo et. al., 2011) that

demonstrate that the differences are only signiﬁcative

after 40/50 years. This is the main limitation of the

present study. Vascular dysfunction of the left and

right carotid arteries is affected by the age, among

other factors. So, left and right carotid signals need

to be further analysed to include different age groups

and correlations with other parameters, such as bio-

chemical parameters.

IndicesandRepeatabilityTestsofCardiovascularFunctionPerformedontheArterialDistensionWaveform-CaseStudy:

AngiographyIntervention

143

5 CONCLUSIONS

The present study investigated the viability of ADW

assessment using a cardiac prototype in a wide num-

ber of subjects, comprising real clinical environment.

The study of different groups during data analysis

demonstrates the ability in different patterns com-

parison for each group. The data variability re-

sults showed good repeatability for the SP, RP and

DN. Therefore, this prototype could be an interest-

ing tool to use in the screening of arterial compli-

cations caused by the arterial stiffness development.

This fact was also proved by the analysis of a stenosis

case study where the clinical alterations were reliably

monitored by our probe.

ACKNOWLEDGEMENTS

Authors acknowledge Fundac¸

ao para a Ci

encia e

Tecnologia for funding SFRH/BD/61356/2009 and

PTDC/SAU-BEB/100650/2008, project developed

under the initiative of QREN, funding by UE/FEDER,

through COMPETE-Programa Operacional Factores

de Competitividade. The authors also thanks to Coim-

bra University Hospital Centre (C.H.U.C.) and Dr.

Rui Provid

encia, SCDSOS-Sudden Cardiac Death

Screening of Risk factors, for the support in clinical

trials.

REFERENCES

Almeida, V. G., Pereira, H. C., Pereira, T., Figueiras, E.,

Borges, E., Cardoso, J. M. R., and Correia, C. (2011a).

Piezoelectric probe for pressure waveform estimation

in ﬂexible tubes and its application to the cardiovas-

cular system. Sensors and Actuators A: Physical,

169(1):217–226.

Almeida, V. G., Santos, P., Figueiras, E., Borges, E.,

Pereira, T., Pereira, H. C., Cardoso, J. M. R., and

Correia, C. (2011b). Hemodynamic features extrac-

tion from a new arterial pressure waveform probe.

Proceedings of the 3rd International Joint Conference

on Biomedical Engineering Systems and Technologies

(BIOSTEC 2011). Rome, Italy.

Avolio, A. P., Butlin, M., and Walsh, A. (2010). Ar-

terial blood pressure measurement and pulse wave

analysis–their role in enhancing cardiovascular as-

sessment. Physiol Meas, 31(1):R1–47. Avolio, Al-

berto P Butlin, Mark Walsh, Andrew England Physiol

Meas. 2010 Jan;31(1):R1-47. Epub 2009 Nov 26.

Bitton, A. and Gaziano, T. A. (2010). The framing-

ham heart study’s impact on global risk assessment.

Prog Cardiovasc Dis, 53(1):68–78. Bitton, Asaf

Gaziano, Thomas A K01 TW007141-05/TW/FIC

NIH HHS/ K01 TW007141-06/TW/FIC NIH HHS/

T32HP10251/PHS HHS/ Prog Cardiovasc Dis. 2010

Jul-Aug;53(1):68-78.

Cavalcante, J. L., Lima, J. A., Redheuil, A., and Al-

Mallah, M. H. (2011). Aortic stiffness: current un-

derstanding and future directions. J Am Coll Cardiol,

57(14):1511–22. Cavalcante, Joao L Lima, Joao A C

Redheuil, Alban Al-Mallah, Mouaz H J Am Coll Car-

diol. 2011 Apr 5;57(14):1511-22.

Conroy, R. (2003). Estimation of ten-year risk of fatal car-

diovascular disease in europe: the score project. Eur

Heart J, 24(11):987–1003.

Hamilton, P. K., Lockhart, C. J., Quinn, C. E., and

McVeigh, G. E. (2007). Arterial stiffness: clini-

cal relevance, measurement and treatment. Clin Sci

(Lond), 113(4):157–70. Hamilton, Paul K Lockhart,

Christopher J Quinn, Cathy E McVeigh, Gary E Eng-

land London, England : 1979 Clin Sci (Lond). 2007

Aug;113(4):157-70.

Huck, C. J., Bronas, U. G., Williamson, E., Draheim, C.,

Duprez, D. A., and Dengel, D. R. (2007). Noninvasive

measurements of arterial stiffness: Repeatability and

interrelationships with endothelial function and arte-

rial morphology measures. Vascular health and risk

management, 3(3):343–349.

Jones, A. F., Walker, J., Jewkes, C., Game, F. L., Bartlett,

W., Marshall, T., and Bayly, G. R. (2001). Compar-

ative accuracy of cardiovascular risk prediction meth-

ods in primary care patients. Heart, 85:37–43.

Laurent, S., e. a. (2006). Expert consensus document on ar-

terial stiffness:methodological issues and clinical ap-

plications. European Heart Journal, 27:25882605.

Liang, Y., Teede, H., Kotsopoulos, D., Shiel, L., Cameron,

J., Dart, A., and McGrath, B. (1998). Non-invasive

measurements of arterial structure and function: re-

peatability, interrelationships and trial sample size.

Clinical Science, 95:669–679.

Luo, X., Yang, Y., Cao, T., and Li, Z. (2011). Differences

in left and right carotid intima-media thickness and

the associated risk factors. Clin Radiol, 66(5):393–8.

Luo, X Yang, Y Cao, T Li, Z England Clin Radiol.

2011 May;66(5):393-8. Epub 2011 Feb 15.

Mackenzie, I. S., Wilkinson, I. B., and Cockroft, J. R.

(2002). Assessment of arterial stiffness in clinical

practice. Q J Med., 95:67–74.

Yusuf, S., Hawken, S., unpuu, S., Dans, T., Avezum, A.,

Lanas, F., McQueen, M., Budaj, A., Pais, P., Varigos,

J., and Lisheng, L. (2004). Effect of potentially mod-

iﬁable risk factors associated with myocardial infarc-

tion in 52 countries (the interheart study): case-control

study. The Lancet, 364(9438):937–952.

BIOSIGNALS2013-InternationalConferenceonBio-inspiredSystemsandSignalProcessing

144