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

net/publication/251176421

Measuring of Systolic Blood Pressure Based On Heart Rate

Article · January 2008

DOI: 10.1007/978-3-540-69139-6_149

CITATIONS READS

4 1,560

4 authors:

Muhamad Khairul Bin Ali Hassan Mohd Yusoff Mashor

Universiti Malaysia Perlis Universiti Malaysia Perlis
12 PUBLICATIONS   12 CITATIONS    178 PUBLICATIONS   1,294 CITATIONS   

SEE PROFILE SEE PROFILE

Nashrul Fazli Mohd Nasir Mohd Sapawi Mohamed

Universiti Malaysia Perlis Hospital Raja Perempuan Zainab II
80 PUBLICATIONS   261 CITATIONS    30 PUBLICATIONS   137 CITATIONS   

SEE PROFILE SEE PROFILE

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

Automated Ki67 Counting for Brain Tumour Grading View project

Wireless Heart Abnormality Monitoring View project

All content following this page was uploaded by Nashrul Fazli Mohd Nasir on 20 March 2014.

The user has requested enhancement of the downloaded file.

 Measuring of Systolic Blood Pressure Based On Heart Rate
M.K. Ali Hassan1, M.Y. Mashor1, N.F. Mohd Nasir2 and S. Mohamed3
1
Mechatronic Engineering Programme,School of Mechatronic Engineering, Universiti Malaysia Perlis,
02600 Jejawi, Arau, Perlis, Malaysia.
2
Biomedical Electronic Engineering Programme,School of Mechatronic Engineering, Universiti Malaysia Perlis,
02600 Jejawi, Arau, Perlis, Malaysia
3
Cardiology Programme,Department of Cardiology, Universiti Sains Malaysia,Kubang Kerian, Kelantan.

Abstract — In this paper, we describe the method of non-
invasive blood pressure (BP) measurement using electrocardi-
ography (ECG) signal to measure arterial blood pressure.
Here, we propose a continuous blood pressure monitoring
method based on heart rates which are more suitable to moni-
tor continuous blood pressure compared to traditional method.
This study examines the relationships between arterial blood
pressure and heart rate from 10 subjects that are normal. The
experiment involved three sessions, which is rest, exercise and
recovery from exercise. The end of this studies is can measured
blood pressure continuously based on heart rate.
Keywords — Blood Pressure, Electrocardiography (ECG),
Heart Rate (HR).
that BP device should have a minimum mean of error read-
out value of 5 mmHg with a standard deviation of error of
±8mmHg. Since SBP is directly proportional to the heart
rate, the regression model using heart rate parameter for
measuring each individual SBP was successful but the re-
gression model using heart rate parameter for estimating
unspecified people’s SBP was not appropriate to our pur-
pose. The current study conducted is to estimate unspeci-
fied individuals’ systolic blood pressure (SBP) by using the
average value of systolic blood pressure from all the sub-
jects and to create a new regression model for each subject.
The flow to create a regression analysis is shown Figure 2.

I. INTRODUCTION

Here, our interest is to propose a continuous blood pres- ECG Signal R-R interval
sure monitoring method based on heart rate. The general
rule is that anything that increases the heart rate increase
blood pressure and anything that decreases the heart rate
causes a decrease in blood pressure. We know that the
T1 T2
blood pressure can be affected by many factors such as
stress, shock, emotion, exercise, sleep and etc. All the fac-
tors that we mentioned above also affect the heart rate. It is
widely accepted that blood pressure will increase with heart
rate and vice versa. This is because the heart rate is directly Figure 1: An illustration of the definition of Heart Rate (HR)
proportional to the cardiac output [1].
The heart rate (HR) can be calculated by dividing the R-
R interval from ECG signal into 60 [2-4]. This number
denoted BPM (Beats per Minute). Electrocardiogram (ECG) ECG Signal Sphygma-
signal are recorded at the human body. Heart Rate can also
be calculated by using pulse transducer where the pulses are
recorded at the fingertip. This study is to estimate systolic
blood pressure continuously and non invasively using ECG Systolic Blood
signal. Heart Rate Pressure

II. METHODOLOGY
Regression Model
Measured each individual systolic blood pressure (SBP)
Figure 2: Flow of regression analysis
and specified people’s SBP based on heart rate, which states

N.A. Abu Osman, F. Ibrahim, W.A.B. Wan Abas, H.S. Abd Rahman, H.N. Ting (Eds.): Biomed 2008, Proceedings 21, pp. 595–598, 2008
www.springerlink.com © Springer-Verlag Berlin Heidelberg 2008
596 M.K. Ali Hassan, M.Y. Mashor, N.F. Mohd Nasir and S. Mohamed

The electrocardiogram (ECG) signals were recorded si- SBP(mmHg) SBP(mmHg) vs HR(BPM)
multaneously for 30 seconds with AD Instruments System. 160

Blood pressure was measured by a standard sphygmoma- 155

150
nometer manually [6-7]. The experimental session involved
145
three sessions: resting period, exercise period and recovery 140 y = 0.7311x + 74.825
period. The first experiment was to estimate systolic blood 135 R2 = 0.8835
pressure for specified people. 130
125
120
1) Regression model based on heart rate. 0 50 HR(BPM) 100 150

Ten healthy subject aged from 19 to 30 years were in-

volved in this experiment. We use AD Instrument System to Figure 4: Data of Subject 2
measure pulse wave transit time. The experiment was con-
ducted as follows SBP(mmHg) SBP(mmHg) vs HR(BPM)
132
y = 0.1659x + 108.18
Step 1) The subject was left to relax for few 130
2
R = 0.8911
minutes. 128

Step 2) Systolic blood pressure (SBP) was measured 126

using sphygmomanometer. 124

Step 3) ECG signal was recorded and the heart rate was 122

measured. 120

Step 4) Step 2 was repeated. 118

0 50 HR(BPM) 100 150
Step 5) Workout is done by the subject for 2 to 3
minutes.
Figure 5: Data of subject 3
Step 6) Step 3 was repeated.
Step 7) Step 2 to 3 was repeated for recovery session.
III. RESULT
Recorded 20 data points for each subjects and used 10
data points to calculate regression model. This data would Table 1 shows the regression models using the heart rate
be verified with other 10 data points and we confirmed its for each subject. It is successful in determining the SBP for
accuracy with sphygmomanometer. The samples of data for personal use but for unspecified peoples it is not suitable.
first three subjects are shown in Figure 3, 4, and 5. The In this section, we apply new method to measure systolic
Heart Rate can be found using the T period from ECG sig- blood pressure for unspecified people. We combine all
nal [8-9]. Here are important equations used: regression models to create a new regression model for each
subject. We averaged the slope values from all regressions
Period, T = t2 – t1 s
Frequency, f = 1 / T Hz and
Heart Rate, BPM = f * 60 s/min. Table 1: Regression model analysis using heart rate
for each individual SBP
Regression Model P-value
SBP(mmHg) SBP(mmHg) vs Heart Rate(BPM) Subject 1 <0.001
134
y = 0.2221x + 107.37
Subject 2 <0.001
132 y = 0.7311x + 74.825
y = 0.2221x + 107.37
130 Subject 3 <0.001
2
R = 0.8762 y = 0.1659x + 108.18
128
Subject 4 <0.001
126 y = 1.1578x + 27.231
124 Subject 5 <0.001
y = 0.633x + 301
122 Subject 6 <0.001
y = 0.1982x + 159.84
120
Subject 7 y = 0.2232x + 168 <0.001
118
0 50 100 150 Subject 8 <0.001
y = 0.5975x + 254.71
Heart Rate(BPM)
Subject 9 <0.001
y = 0.6946x + 251.71
Figure 3: Data of Subject 1 Subject 10 <0.001
y = 0.3254x + 362.1

_________________________________________ ___________________________________________
IFMBE Proceedings Vol. 21
Measuring of Systolic Blood Pressure Based On Heart Rate 597

Table 2: New Regression model Table 4: The new SBP data of Subject 2 after applying new method
Regression Model New Regression Model HR (BPM) SBP New SBP (mmHg) Different (mmHg)
Subject 1 y = 0.2221x + 107.37 y = -0.41482x + 256.15 (mmHg)
Subject 2 y = 0.7311x + 74.825 y = -0.41482x + 159.15 76.92308 128 128 0
Subject 3 y = 0.1659x + 108.18 y = -0.41482x + 233.58 75.94937 127 127.4457 0.4457
Subject 4 y = 1.1578x + 27.231 y = -0.41482x + 191.91 73.17073 128 125.8641 -2.13588
Subject 5 y = 0.633x + 301 y = -0.41482x + 285 74.07407 128 126.3783 -1.6217
Subject 6 y = 0.1982x + 159.84 y = -0.41482x + 171.22 107.5269 150 145.4202 -4.5798
Subject 7 y = 0.2232x + 168 y = -0.41482x + 237.28 105.4482 150 144.2369 -5.7630
Subject 8 y = 0.5975x + 254.71 y = -0.41482x + 178.66 90.63444 146 135.8047 -10.1952
Subject 9 y = 0.6946x + 251.71 y = -0.41482x + 259.66 94.93671 146 138.2536 -7.7463
Subject 10 y = 0.3254x + 362.1 y = -0.41482x + 198.975 86.95652 142 133.7112 -8.2888
85.59201 142 132.9345 -9.0655

model. Afterward, a new slope would be used as a reference

for all the regression models. New regression model for all Table 5: The new SBP data of Subject 1 after applying new method
subjects are shown in Table 2. HR (BPM) SBP New SBP (mmHg) Different (mmHg)
After applying new method, the result is exhibited in Ta- (mmHg)
ble 3, Table 4 and Table 5 exhibit both data origin from
69.04488 108 108 0
measurement using sphygmomanometer and new method.
74.07407 108 110.8627 2.8626
The value of estimated SBP is near to the SBP value using
68.88634 105 107.9098 2.9097
traditional method.
72.02881 105 109.6985 4.6985
From Table 3, 4 and 5 the estimated value and measured
83.91608 130 117.4649 -12.5350
value for normal blood pressure (91 mmHg to 129mmHg) is
84.03361 130 117.5318 -12.4681
very close. The different value of estimated value and
83.10249 120 116.0018 -3.9981
measured value is small. Now we can continue to measure
SBP value for other subjects using this method. The step to 81.96721 120 115.3556 -4.6444
measure new SBP as follows 66.15215 111 106.3534 -4.6465
71.42857 111 109.3568 -1.6431
Step 1) SBP was measured using sphygmomanometer
Step 2) Heart rate value was measured from ECG According to Table 3, 4 and 5 the regression model based
signals. on heart rate parameters use to estimate unspecified peo-
Step 3) New regression model was generated ple’s SBP is appropriate and accurate after we checked the
Step4) The regression model is used to measure accuracy with sphygmomanometer.
SBP continuously onward.

Table 3: The new SBP data of Subject 1 after applying new method IV. CONCLUSION
HR (BPM) SBP New SBP (mmHg) Different (mmHg) After applying new method which origin from average
(mmHg)
value of slope from the regression model, the unspecified
66.96429 121 121 0 people’s SBP estimation is appropriate. Although the re-
73.08161 122 124.4821 2.4820 gression model based on heart rate gives acceptable results,
69.93007 123 122.6882 0.3118 it requires more study to make it more accurate and robust.
58.82353 123 116.3662 6.6338 Our future plan is to expand the study to individuals with
105.6338 132 143.0113 11.0113 the history of high blood pressure.
105.6338 132 143.0113 11.0113
94.78673 128 136.837 8.8369
94.78673 127 136.837 9.8369 ACKNOWLEDGMENTS
73.3126 123 124.4821 1.4821
70.42254 124 122.9685 1.0315 This study was supported by the University Malaysia
Perlis Fund, grant no. 9005-00024.

_________________________________________ ___________________________________________
IFMBE Proceedings Vol. 21
598 M.K. Ali Hassan, M.Y. Mashor, N.F. Mohd Nasir and S. Mohamed

REFERENCES 7. D.Perloff, M. Sokolow and R. Cowan, “The prognostic value of

ambulatory blood pressure, JAMA, pp. 240: 1308 – 1309, 1983
1. I. G. Webster, Medical Instrumentation: Application and design, third 8. M. Malik, and A. J. Camm, Heart Rate Variability. Armonk,NY:
edition, New York: Wiley, 1998. Futura, 1995
2. Blood Pressure Monitoring, August 2006, 9. Task Force of the European Society of Cardiology, the North Ameri-
http://www.medphys.ucl.ac.uk can Society of Pacing, and Electro physiology, “ Heart Rate Variabil-
3. V. X. Afonso, W.J. Tompkins, “ECG Beat Detection Using Filter ity: Standards of measurement, physiology interpretation, and clinical
Banks,” IEEE Transactions on Biomedical Engineering, vol. 46, use”, Sofia Antipolis, France, vol.93, 1996
n0.2,pp. 192-202, February 1999
4. P.Chazal, M.O’Dwyer, and R. B. Reilly, “Automatic Classification of
Heartbeats Using ECG Morphology and Heartbeat Interval Features,”
IEEE Transactions on Biomedical Engineering, vol. 1, no.7, pp. 1196-
1206, July 2004
5. R. Silipo, and C. Marchesi, “Artificial neural network for automatic Author: Muhamad Khairul Bin Ali Hassan
ECG analysis”, IEEE Transaction on Signal Processing vol. 46, no. 5, Institute: Universiti Malaysia Perlis
pp. 1417-1425, May 1998 Street:
6. G. Mancia, A. Ferrari, and L.Gregorini “Relationship between level City: Kangar, Perlis
of blood pressure measured casually and by portable recorders and Country: Malaysia
severity of complications in essential hypertension.” pp. 279-298, Email: muhamad_khairul@yahoo.com
1966

_________________________________________ ___________________________________________
IFMBE Proceedings Vol. 21

View publication stats