ECG SIGNAL SYNTHESIS FOR HEART DISEASES

ABSTRACT

An electrocardiogram (ECG), which is the electrical activity of the heart is used

by medical experts to diagnose various heart diseases. Variations in the heart
rate, amplitude and duration of P wave, QRS Complex and T wave are
considered as indications of heart diseases. Cardiovascular disease has
emerged as a major threat to human health, and the number of patients
suffering from it is increasing. Experienced doctors find it easy to visually
observe the ECG signal and find out the minute changes in it. While new
doctors find it difficult to identify these minute changes. So, there is a need for
standard ECG for different diseases so that doctors can compare the measured
ECG with this. The objective of this project is to synthesise ECG signals for
different heart diseases. In the later phase, deep learning will be used to
generate signals according to the diseases. The imbalance between a shortage
of medical resources and an increase in the number of patients has become
increasingly apparent in this environment. Since computer-assisted diagnosis
of cardiovascular disease has become increasingly common, the study of the
ECG automatic classification approach has significant practical implications.

Guided by-

Dr. Sinith M S
Assistant Professor
ECE Department
GEC Thrissur
\

The P wave, QRS complex, and T wave, when combined, form a normal ECG
wave. Cardiovascular diseases are indicated by deviations from this
normal graphical representation.
INTRODUCTION

An electrocardiogram (ECG) is a standard test for determining and diagnosing

the working condition of the heart, in which the heart's rhythm and activity are
recorded on a moving strip of paper or presented as a line on the screen of an
ECG Oscilloscope. The doctor can then interpret the data and study the peaks
and dips to look for any abnormal or irregular heart activity. It is necessary to
understand the characteristics, and appearance of such ECG waveforms in
various medical institutions and laboratories to do the above-mentioned
analysis.

The P wave, QRS complex, and T wave, when combined, form a normal ECG
wave. It has several peaks and troughs. Cardiovascular diseases are indicated
by deviations from this normal graphical representation.

The cardiac conduction system is a collection of nodesband specialized

onducion cells that initiate and coordinate contraction of heart muscle. The
ECG signal is generated by depolarization and repolarization of the atrial and
ventricular chambers. This depolarization and repolarization are initiated by
an electrical signal produced by the sinoatrial node (SA node) in the heart. This
signal reaches the Atrio-Ventricular Node ( AV node) which electrically
connects atria and ventricles. The bundle of HIS, which is a collection of heart
muscle cells carry electrical impulses from the AV node to the bundle
branches. The signal from the bundle of HIS splits into left and right bundle
branches, directing the impulse to left and right ventricles respectively.
Perkinje fibers are the most distal component of the cardiac conduction
system.

The heart signal is called a PQRST wave which has many peaks and troughs.
Atrial and ventricular depolarization and repolarization are responsible for the
generation of the ECG signal. The sinoatrial node in the heart creates a small
electrical signal which is responsible for this depolarization and repolarization.
ECG is represented as a series of waves: the P wave followed by the QRS
complex and the T wave. The J point is found at the junction of the QRS
complex and the beginning of the ST segment.
The statistics which provides the estimation of cardiovascular mortality in India
are more than 4.5 million deaths in 2018 at all ages.
About 3.1 million cardiovascular deaths are estimated accounts for people
between 30-69 years. Coronary heart disease was found in over 1.5 million
(68.4%) of the total population and stroke for about 1 million (28.0%) people.
Due to improper diagnosis, 15% of the total population die from cardiac arrest.
The doctors must compare the patient's ECG with the reference to identify the
diseases. Hence, the need for the synthetic ECG generator plays a major role.

Incorrect diagnosis leads to the death of 15% of the total population. Hence,
there is a need for the synthesis of ECG signals so that they can be used as a
reference to identify cardiovascular diseases.

Over 75% of CVD deaths take place in low- and middle-income countries
where raised blood pressure happens to be amongst the most important
risk factors for CVDs. In 2016 India reported 63% of total deaths due to
NCDs, of which 27% were attributed to CVDs. CVDs also account for 45%
of deaths in the 40-69 year age group.

In 2016, 17.9 million individuals died from cardiovascular disease (CVD),

accounting for 31% of all deaths worldwide based on the studies conducted
by the World Health Organization (WHO). . Heart attacks and strokes were
responsible for about 80 percent of these deaths.

In India, chronic noncommunicable diseases (NCDs) were responsible for

63 percent of all fatalities, with CVDs accounting for 27 percent. CVDs are
also responsible with 45 percent of mortality in people aged 40 to 69.

According to the Global Burden of Disease, CVDs account for over a

quarter (24.8%) of all fatalities in India.
Even a review of medical certification of cause of death (MCCD) data
shows that the proportion of deaths caused by CVD is rising.

Incorrect diagnosis leads to the death of 15% of the total population. But there
is no standard reference to identify diseases. Hence, there is a need for the
synthesis of ECG signals so that they can be used as a reference to identify
cardiovascular diseases. It can also be used to calibrate an ECG recorder.

SOFTWARE IMPLEMENTATION4.1
MATLAB
MATLAB (matrix laboratory) is a multi-paradigm programming language and numeric computing
environment developed by MathWorks. It allows matrix manipulations, plotting of functions and
data, implementation of algorithms, creation of user interfaces, and interfacing with programs
written in other languages. MATLAB help to analyze signals using built-in apps for visualizing and
preprocessing signals in time, frequency, and time-frequency domains to detect patterns and trends
and for algorithm development. Signal Processing Toolbox provides functions and apps to analyze,
preprocess, and extract features from uniformly and non uniformly sampled signals. The toolbox
includes tools for filter design and analysis, resampling, smoothing, detrending, and power spectrum
estimation. Filter Designer is a powerful graphical user interface (GUI) in Signal Processing Toolbox
for designing and analyzing filters. Filter Designer enables to quickly design digital FIR or IIR filters by
setting filter performance specifications, by importing filters from the MATLAB workspace, or by
adding, moving, or deleting poles and zeros. Filter Designer also provides tools for analyzing filters,
such as magnitude and phase response plots and pole-zero plots.
The heart is the main organ of the cardiovascular system. It is primarily
responsible for pumping blood and distributing oxygen and nutrients throughout
the body. Even small abnormalities or dysfunctions may cause drastic effects.
The heart is divided into four chambers- two upper chambers known as the
atrium and two lower chambers known as ventricles. During joint diastole (0.4s),
all the chambers are relaxed. De-oxygenated blood from different body parts
enters the right auricle through the superior and inferior vena cava. Blood enters
the right ventricle due to auricle systole (0.1s). Due to ventricle systole (0.3s), it
moves to the lungs for gas exchange through the pulmonary artery. Oxygenated
blood from the lungs reaches the left auricle through the pulmonary vein. Due to
auricle systole, it enters the left ventricle. Due to ventricle systole, it is moved to
body parts through the aorta. This constitutes one cardiac cycle with a total
duration of 0.8 seconds.
LITERATURE REVIEW
In [1], the major objective of the project was to synthesize and analyze ECG
signals using Fourier transform and wavelet transform and compare their
performance. Fourier transform analyzes a signal only in the frequency domain.
As a result, the time localization information is lost. Hann window technique is
was used for extracting more information using Fourier transform. But this
method failed to reconstruct the signal effectively. To overcome this drawback,
wavelet transform (Continous and discrete) was used which will provide both
time-domain and frequency domain information. The authors concluded that
wavelet transform is an effective clinical tool to accurately detect heart diseases
compared to the Fourier transform.

An ECG signal generator was developed in [2] to help doctors to analyze

ECG signals. This generator can also generate ECG signals for various diseases by
changing some values. A synthetic ECG signal was generated using Fourier series
analysis to be used as a preliminary wave. ECG waveform is generated by
repeating combinations of sine waves. It was also used to generate the signal for
various types of heart diseases including myocardial infarction and hyperkalemia.

In [3], the objective of the authors was to develop a mathematical model for
generating synthetic artificial ECG signals. Modelling is done using trigonometric
function and Gaussian monopulse. By proper addition of respiration frequency
and equipment noise, this artificial signal can be used in signal processing
methods as a test signal.

An automatic system involving MATLAB tool was used to synthesise ECG signals in
[4]. The mathematical modelling is done using the Fourier series. The features
extracted from real ECG data was used to shape the model.
REFERENCES

[1] M. Basheer and I. Janajreh, “Analysis and Synthesis of Electrocardiogram (ECG)

using Fourier and Wavelet Transform,” Int. J. Therm. Environ. Eng., vol. 17, no. 02,
pp. 1–12, 2020, doi: 10.5383/ijtee.17.02.002.

[2] A. kumar Srimugi ,Kishore Balasubramanian, Shurthi, “Synthetic Generation of

ECG Signal for the Analysis of Heart Diseases,” Int. J. Res. Anal. Rev., vol. 7, no. 1,
pp. 148–154, 2020.

[3] B. R. Manju and B. Akshaya, “Simulation of Pathological ECG Signal Using

Transform Method,” Procedia Comput. Sci., vol. 171, no. 2019, pp. 2121–2127,
2020, doi: 10.1016/j.procs.2020.04.229.

[4] B. R. Manju and B. Akshaya, “Simulation of Pathological ECG Signal Using

Transform Method,” Procedia Comput. Sci., vol. 171, no. 2019, pp. 2121–2127,
2020, doi: 10.1016/j.procs.2020.04.229.

[5] J. Sohn, S. Yang, J. Lee, Y. Ku, and H. C. Kim, “Reconstruction of 12-lead

electrocardiogram from a three-lead patch-type device using a LSTM network,”
Sensors (Switzerland), vol. 20, no. 11, pp. 1–13, 2020, doi: 10.3390/s20113278.
(