Wavelet based ecg signal component identification

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
_______________________________________________________________________________________
Volume: 03 Issue: 05 | May-2014, Available @ http://www.ijret.org 703
WAVELET BASED ECG SIGNAL COMPONENT IDENTIFICATION
Tatiparti Padma1
, Chepuri Bhavana2
, Peddamallu Sridivya3
1
Faculty, Department of Biomedical Engineering, GRIET, AP, INDIA
2
Student, Department of Biomedical Engineering, GRIET, AP, INDIA
3
Student, Department of Biomedical Engineering, GRIET, AP, INDIA
Abstract
The most commonly used technique to record the bioelectric current generated by the heart is Electrocardiogram. It is a simple,
painless and valuable test done by placing the surface electrodes at the predetermined locations on the surface of the heart. It is a
composite signal consisting of P, QRS, T and U waves, vital information related to arrhythmias (heart rhythm abnormalities) can
be extracted from intervals and segments of these waves. The present study includes the development of an algorithm which pre-
processes and automatically extracts the features of an ECG signal using wavelet transformation. As Wavelets provide both
temporal and spatial information simultaneously and offers flexibility with the choice of wavelet functions with different
properties. The initial step of the algorithm includes pre-processing of the signal in order to remove noise from the signal
(denoising) and then uses the pre-processed signal for feature extraction. The accuracy of the analysis is increased and also the
analysis time is reduced by using this algorithm.
Keywords— ECG, P wave, QRS complex, T wave, R-R interval, ST interval, QT interval, ST segment, heart rate,
Wavelet transform.
---------------------------------------------------------------------***---------------------------------------------------------------------
1. INTRODUCTION
The action potentials generated by all the heart muscle fibers
during each heart beat collectively forms ECG record.
Electrocardiograph is an instrument used to record the
changes in ECG. During diagnosis, electrodes are positioned
on the arms and legs (limb leads) and at six positions on the
chest (chest leads). The electrocardiograph is capable of
amplifying the heart’s electrical signals and produces 12
different tracings from various combinations of limb and
chest leads. The slight difference in the electrical activity is
due to placement of limb and chest electrodes at different
positions relative to the heart. By comparing these records
with one another and with normal records, it is possible to
determine (i) if the conducting pathway is abnormal, (ii) if
the heart is enlarged, (iii) if certain regions of the heart are
damaged, and (iv) the cause of angina. In a typical record,
three distinct waves appear with each heartbeat. The first
wave is P wave, it is a small upward deflection on the ECG.
The P wave represents atrial depolarization. The second wave,
is QRS complex, begins as a downward deflection, continues
as a large, upright, triangular wave, and ends as a downward
wave. The QRS complex represents rapid ventricular
depolarization, as the action potential propagates through
ventricular contractile fibers. The third wave is a dome-
shaped upward deflection called the T wave. It indicates
ventricular repolarization. T wave occurs during the
relaxation phase of ventricles. Though it is not always visible,
U wave is considered as the representation of Papillary
Muscle or Purkinje Fibers [2]. ECG waveform characteristics
and their corresponding positions in heart is shown in the
Fig.1 given below. As repolarisation is much slower than
depolarization, the T wave is smaller and wider than the QRS
complex. Based on Electrocardiogram signals many of the
heart rhythm abnormalities (arrhythmias) can be diagnosed,
during this process a physician analyses the signal and
extracts the features like amplitudes of the waves and the
time interval between them. Larger P waves indicate
enlargement of an atrium, an enlarged Q wave may indicate a
myocardial infarction and an enlarged R wave generally
referred as enlarged ventricles [3]. Due to insufficiency of
oxygen, the T wave appears flatter than normal T wave.
Fig. 1 ECG waveform characteristics and their corresponding
positions in heart [1]
For analysis of an ECG intervals or segments (measurement
of time between two waves) are to be considered. For
example, The Q-T interval extends from the start of the QRS
complex to the end of the T wave. It is the duration from
beginning of ventricular depolarization to the end of
ventricular repolarization. The Q-T interval may be
lengthened by myocardial damage, myocardial ischemia
(decreased blood flow), or conduction abnormalities. The S-T

_______________________________________________________________________________________
interval is from end of S to end of T-wave. The normal
values of the ECG signal are shown in the Table I below and
the typical normal ECG signal is shown in Fig.2. Any change
in the given listed values indicates the abnormality of the
heart [3].
Table 1
S.No Features Amplitude/Duration
1 P wave 0.25 mV
2 R wave 1.6mV
3 Q-wave 0.4mV
4 T-wave 0.1 to 0.5 mV
5 R-R interval 0.6-1.2 Sec
6 PR-interval 0.12-0.2 Sec
7 QRS complex 0.04 to 0.12Sec
8 S-T interval <0.42 Sec
9 QT interval <0.44 Sec
10
Normal Heart
Rate
60-100 BPM
Fig. 2 Normal ECG signal [4]
2. WAVELET TRANSFORMS
The mathematical tool for decomposing a signal into a set of
orthogonal waveforms both in time and frequency domains is
wavelet transform. Wavelets are functions that are used to
represent data or other functions and satisfy certain
mathematical requirements. In wavelet analysis, the scale
used to look at data plays a special role. Morlet first
considered wavelets as a family of functions constructed
from translations and dilations of a single function called the
"mother wavelet" ψ(t). They are defined by
ψa,b 𝑡 =
1
𝑎
ψ
t−b
a
, a,b ∈ R,a ≠ 0 (I)
‘a’ is the scaling parameter or scale, and it measures the
degree of compression and b is the translation parameter
which determines the time location of the wavelet. When |a|
< 1, then the wavelet is in the compressed version (smaller
support in time domain) of the mother wavelet and
corresponds mainly to higher frequencies. On the other hand,
when |a| > 1, then ψa,b(t) has a larger time-width than ψ(t) and
corresponds to lower frequencies. Thus, wavelets have time-
widths adapted to their frequencies. This is the main reason
for the success of the Morlet wavelets in signal processing
and time-frequency signal analysis. Data can be processed at
different scales or resolutions using Wavelet algorithms.
Fig. 3 2nd order DWT decomposition and reconstruction
structure [5]
The importance of wavelet analysis is to adopt a wavelet
prototype function, called an analysing wavelet or mother
wavelet. In the wavelet analysis the filter decomposes the
signal into frequency bands. The decomposed signal can be
reconstructed to the original bands using filter in wavelet
synthesis. Both the shape and perfect reconstruction of
wavelet can be determined by selection of filter. Discrete
wavelet transform has number of advantages when applied
towards ECG analysis. Discrete Wavelet Transform (DWT)
has two filters; a Low Pass Filter (LPF) and a High Pass
Filter (HPF).They are used to decompose the signal into
different scales. The output coefficients of the LPF are called
‘Approximation’ while the output coefficients of the HPF are
called ‘Detail’. The Approximation signal can be sent again
to the LPF and HPF of the next level for second-level
decomposition, thus to decompose the signal into its different
components at different scale levels [5]. Decomposition and
reconstruction structure of wavelet is shown in Fig. 3.
Some of the wavelet families are Haar, Daubechies, symlets,
coiflets etc. Commonly used wavelets for ECG signal
processing is Daubechies. In this study, an algorithm is
developed to extract the ECG features automatically. The
developed algorithm initially decomposes the original ECG
signal by using Wavelet Transform and Daubechies wavelet
(db4) as mother wavelet. Wavelets are a powerful statistical
tool which can be used for a wide range of applications,
namely Signal processing, DNA analysis, protein analysis,
Speech recognition. Wavelets have been used successfully in
other areas of geophysical study.

_______________________________________________________________________________________
3. FEATURE IDENTIFICATION USING
WAVELETS
The sample ECG signals for the present study are collected
from MIT-BIH database from Physionet website and stored
in a text format. The MIT-BIH database contains many types
of ECG signals including both normal and abnormal, which
are sampled at different rates [6]. For example record number
16272 is originally sampled at 128 Hz, record 30 was
sampled at 250 Hz and record 113 was sampled at 360 Hz.
Therefore, to process all the signals uniquely, all the samples
must be resampled at 360 Hz before processing the ECG
signal. Each selected ECG signal is of thirty minute duration,
but only 5 minutes duration of the signal is used for
processing in this study.
Discrete Wavelet transform are used to extract features from
an electrocardiogram (ECG). Detection can be achieved in a
better way by the wavelet filter with scaling function similar
to the shape of the ECG signal. The wavelet is similar in
shape to QRS complex and their energy spectrum is
concentrated around low frequencies excluding the high
frequency component and is an approximated ECG signal.
Some of the components for analysis of an ECG signal are
listed below.
3.1 P and T-Wave
Function of atrium and Sino-atrium node (SA) will detect the
shape and the time interval of P-wave. It intern will provide
the pace to the following components of ECG complex. Thus,
a theoretical link between frequency domain spectral analysis
techniques and time domain analysis is constructed. The T
wave is the positive deflection after each QRS complex. It
represents ventricular repolarization. The T wave contains
more information than the QT interval.
3.2 QRS Complex
Most of the energy is concentrated in QRS complex of ECG
waveform. It provides reasonably good amount of
information about the functioning capability and the pacing
ability of human heart. The variation of amplitude or/and
time of QRS complex can help in determining the symptoms
required for study. QRS interval or duration extends from
beginning of Q to the end of the S wave. If QRS is greater
than 0.12 sec, indicates the presence of bundle branch block.
3.3 ST Segment
ST segment begins at the end of the S wave and ends at the
beginning of the T wave. During the plateau phase of the
action potential, the ventricular contractile fibers are
depolarized that represents the ST segment. In acute
myocardial infarction the ST segment is elevated (above the
baseline) and depressed (below the baseline) when the heart
muscle receives insufficient oxygen.
3.4 QT, ST and RR Intervals
The QT interval extends from the beginning of the QRS
complex to the end of the T wave. It is the time duration
between ventricular depolarization to ventricular
repolarization. The duration of QT interval exceeds (i) in
myocardial damage, (ii) myocardial ischemia (decreased
blood flow), or conduction abnormalities and (iii)
Hypothyroidism. The ST interval is measured from the J
point to the end of the T wave. Junction between the end of
QRS and beginning of ST segment is called J point (used to
measure degree of ST elevation or depression present). The
duration between two consecutive R waves represents RR
interval and its normal range is 0.6 to 1.2 sec.
3.5 Heart Rate
R Wave is the most dominant wave in ECG complex. The
synchronization between the two pace making nodes can be
reflected with the normal R-R interval. The heart rate can be
calculated by successive averaging of time duration between
two consecutive R waves and it is measured in Beats Per
Minute (BPM) [1].
4. PROCEDURE FOR THE PROPOSED
ANALYSIS
Fig.4 Flow chart for features extraction of ECG signal
The acquired ECG signal is preprocessed to remove noise
due to artifacts caused by motion, power line interferences,
electrode contact noise, muscle contractions and breath
caused baseline movement by decomposing the signal using
daubechies (db4) wavelet. The resultant or preprocessed
signal is then used to detect all the peaks and then followed
by the extraction of above listed parameters. Fig. 4 reflects
the flowchart of the developed algorithm.

_______________________________________________________________________________________
5. EXPERIMENTAL RESULTS
The algorithm is implemented on various signals of MIT-
BIH database. Here 2000 samples of every signal is
considered for experimentation. Following are results for the
signal 103 is shown in Fig 5(a) and (b).
Fig 5(a) Original signal
Fig.5 (b) Detected all peaks in ECG signal
From the above detected peaks, the values of ST interval-
0.0108sec, QT interval-0.0076sec, RR interval- 0.85 sec and
heart rate is 71(normal), QRS duration-0.0254 sec.
6. CONCLUSIONS AND FUTURE WORK
The study mainly involves the development of an algorithm
that extract the features more accurately. One of the
important steps in the ECG analysis is to detect the different
waves forming the entire cardiac cycle. Most of the studies
based around wavelet transformation identify 99.8% of ECG
waveforms. The wavelet transformation is worth enough in
recognition of P and T wave.
Wavelet transformation in electro cardiology is relatively
new field of research. Many methodological aspects (Choice
of the mother wavelet, values of the scale parameters) of the
wavelet technique will require further investigations in order
to improve the clinical usefulness of this novel signal
processing technique. The extracted features can be used
further for classification of different heart rhythm
abnormalities like tachycardia, Bradycardia and myocardial
ischemia etc.
REFERENCES
[1] J. L. Rajput and K. C S, "Classification of ECG
abnormalities using ANN," International Jounral of
Research and Technology, vol. 1, no. 5, 2012
[2] B. D. Gerard J. tortora, Principles of Anatomy and
Physiology
[3] V. S. Y. N. M. C.Saritha, "ECG Signal Analysis
Using Wavelet Transforms," Bulg. J. Phys., vol. 35,
pp. 68-77, 2008
[4] C. Guyton and J. E. Hall, Text bok of Medical
Physiology, Philadelphia: Elsevier Inc, 2006
[5] R. G. Rasika Burte, "Advances In QRS detection :
Modified Wavelet Energy Gradient Method,"
International Journal of Emerging Trends in Signal
Processing, vol. I, no. I, p. 23, 2012
[6] "Physionet.org,"[Online]. Available:
http://www.physionet.org/physiobank/database/mitdb/
index.html

Wavelet based ecg signal component identification

More Related Content

What's hot (18)

Viewers also liked (20)

Similar to Wavelet based ecg signal component identification (20)

More from eSAT Publishing House (20)

Recently uploaded (20)

Wavelet based ecg signal component identification