Efficient data compression of ecg signal using discrete wavelet transform

IJRET: International Journal of Research in Engineering and Technology ISSN: 2319-1163
__________________________________________________________________________________________
Volume: 02 Issue: 04 | Apr-2013, Available @ http://www.ijret.org 696
EFFICIENT DATA COMPRESSION OF ECG SIGNAL USING
DISCRETE WAVELET TRANSFORM
Karishma Qureshi1
, V. P. Patel2
1
PG Student, 2
Head of the Department, Applied Instrumentation, L. D. College of Engineering, Gujarat, India
qureshikarishma@gmail.com, vinodpatel_74@rediffmail.com
Abstract
Data compression reduces the number of bits of information required to store or transmit the biomedical signals. Compression
algorithm of bio-medical signals is implemented using Discrete Wavelet Transform. High threshold value (λ) gives high data
reduction and poor signal fidelity and low threshold value (λ) gives low data reduction and high signal fidelity. Threshold value
selection should be such that the quality of the ECG signal is not distorted on reconstruction and a good amount of data reduction is
also achieved. The database has been collected from MIT-BIH arrhythmias database of the lead II (ML II) signal. The ECG signal to
be compressed is decomposed to the Level 5 using the biorthogonal 4.4 wavelet family. In this paper different issues for compression
are demonstrated and the results are shown using MATLAB
Index Terms: Electrocardiogram (ECG), Discrete Wavelet Transform (DWT), Compression Ratio (CR), Percentage Root
Mean Deviation (PRD), etc.
-----------------------------------------------------------------------***-----------------------------------------------------------------------
1. INTRODUCTION
There is an exponential increase in digital data, obtained from
various signals specially the biomedical signals such as
electrocardiogram (ECG), electroencephalogram (EEG),
electromyogram (EMG) etc. How to transmit or store these
signals efficiently becomes the most important issue. [1]
Transmission techniques of these biomedical signals through
communication channels is an important issue as it allows
experts to make a remote assessment of the information
carried by the signals, in a very cost-effective way. Storage of
these bio-medical signals leads to a large volume of
information. Thus the necessity of efficient data compression
methods for biomedical signals is currently widely recognized.
[3]
The compression of Electrocardiogram signals is done using
Discrete Wavelet Transform (DWT). Thus, a digital
compression technique is used to solve this problem. In this
paper, the compression of ECG signal has been done with
biorthogonal wavelet function. The parameters such as
compression ratio and percentage root mean deviation are
calculated for different samples of ECG signal. 1024, 2048,
4096, 8192, 16384 samples are taken for record 100 from
MIT-BIH arrhythmias database and compression is done.
2. ECG SIGNAL COMPRESSSION
ECG compression techniques can be categorized into: direct
time-domain techniques, transformed frequency domain
techniques and parameters optimization techniques.[3]
2.1 Direct Signal Compression Techniques
A direct method performs the compression directly on the
ECG signal. These are also known as time domain techniques.
To get a high performance time domain compression
algorithm, intelligent sample selection criteria should be used.
The original signal is reconstructed by an inverse process,
often by drawing straight lines between the extracted samples.
The key to a successful algorithm is the development of a
good rule for determining the most significant samples. [3]
2.2 Transformed ECG Compression Methods
Transform method, converts the time domain signal to the
frequency or other domains and analyzes the energy
distribution. These methods mainly utilize the spectral and
energy distributions of the signal by means of some transform,
and properly encoding the transformed output.
Signal reconstruction is achieved by an inverse transformation
process. This category includes different transform techniques
such as the Karhunen–Loève transform, Fourier transform,
Cosine transform, sub band-techniques, vector quantization
(VQ), and more recently the wavelet transform. Wavelet
technique is the obvious choice for ECG signal compression
because of its localized and non-stationary property and the
well-proven ability of wavelets to see through signals at
different resolutions. The main task in wavelet analysis
(decomposition and reconstruction) is to find a good analyzing
function (mother wavelet) to perform an optimal
decomposition. [3]

__________________________________________________________________________________________
2.3 Optimization Methods for ECG Compression
Optimization method minimizes the reconstruction error given
a bound on the number of samples to be extracted or the
quality of the reconstructed signal to be achieved. The ECG
signal is compressed by extracting the signal samples that,
after interpolation, will best represent the original signal given
an upper bound on their number. The aim is to present ECG
compression technique that achieves maximum data volume
reduction while preserving the significant signal morphology
features upon reconstruction.[3]
3. DIFFERENT TRANSFORM TECHNIQUES
3.1 Discrete Fourier Transform:[1]
Discrete Fourier Transform is a fundamental transform in
digital signal processing with applications in frequency
analysis, signal analysis etc.
The uth DFT coefficient of length N sequence {f(x)} is
defined as in Equation 3.1: [1]
….…. Equation 3.1
Where u = 0, 1,…..…….N-1
Whereas its inverse transform (IDFT) as in Equation 3.2:
……Equation 3.2
Where x = 0, 1,…..…….N-1
The number of complex multiplications and additions to
compute DFT is N2.
3.2 Discrete Cosine Transform:[1]
The Discrete Cosine Transform is closely related to DFT but
only uses only real numbers, it express a function or a signal
in terms of a sum of sinusoid with different frequencies and
amplitudes. DCT is often used in signal and image processing,
especially for lossy data compression, because it has a strong
"energy compaction" property .Signal can be reconstructed
fairly accurately from only few DCT coefficients.
A discrete Cosine Transform of N sample is defined as in
equation 3.3: [1]
Equation 3.3
Where u = 0, 1,…..…….N-1
The inverse DCT is shown in equation 3.4: [1]
Equation 3.4
Where x = 0, 1,…..…….N-1
3.3 Discrete Wavelet Transform:[1]
Wavelets allow both time and frequency analysis of signals
simultaneously because of the fact that energy of wavelet is
concentrated in time and still possesses the wave like
characteristics. Therefore it is suitable for the analysis of time-
varying non-stationary signals such as ECG.
Wavelets are functions generated by one single basis function,
called mother wavelet ψ (t) by dilation ‗a‘ and translation ‗b‘
represented as in equation 3.5:[1]
………...Equation 3.5
Where a = scaling parameter and b= shifting parameter
4 DIFFERENT COMPRESSION PARAMETERS
4.1 Compression Ratio (CR):
The compression ratio (CR) is defined as the ratio of the
number of bits representing the original signal to the number
of bits required to store the compressed signal. All data
compression algorithm, used to minimize data storage by
eliminating the redundancy wherever possible to increase the
compression ratio. Compressed data must also represent the
data with better fidelity while achieving high compression
ratio [2].
Compression Ratio (CR)
A high compression ratio is typically desired. [2]
4.2 Percentage Root Mean Difference (PRD):
Percentage root mean difference (PRD) measures distortion
between the original signal and the reconstructed signal.

__________________________________________________________________________________________
PRD can be defined as:
PRD = * 100
Where x1(n) = original signal of length N and
x2(n) = reconstructed signal of length N
The PRD which is an error measurement indicates
reconstruction fidelity by point wise comparison with the
original data. [2]
5 COMPRESSION ALGORITHM:[2]
Step 1: Loading of ECG signal from MIT-BIH arrhythmia
data base (record 100)
Step 2: Transform the original ECG signal using DWT
Step 3: To achieve an adaptive threshold compute the
maximum value of the transformed coefficients.
Step 4: Apply the threshold of a fix percentage based on
absolute maximum values of the transform coefficients.
Step 5: Apply inverse transform to get the reconstruct signal.
Step 6: Calculation of Percent Root Mean Square Difference
(PRD).
Step 7: Calculation of Compression Ratio (CR).
6 DATA COLLECTION
Loading the ECG signal. Database collection is the one of the
most important part of signal processing. MIT-BIH arrhythmia
Database directory is selected from physionet. The ECG
signals are sampled at 360 Hz (i.e. 360 samples/second) and
each sample is represented by 11 bits/sample. In this work,
record 100 of duration 1 minute is taken i.e. 1 second gives
360 samples, 10 seconds gives 3600 samples and 60 seconds
(1 minute) gives 21600 samples. Compression algorithm is
tested for 1024, 2048, 4096, 8192, 16384 samples of record
100 from MIT-BIH arrhythmia Database.
7 RESULTS
Figure 1 (a) shows the original ECG signal and (b) shows the
reconstructed signal for 1024 samples of record 100 from
MIT-BIH arrhythmia database with Threshold λ=280.7,
Compression Ratio = 34, Percentage Root Mean Deviation =
2.37%
0 200 400 600 800 1000 1200
850
900
950
1000
1050
1100
1150
1200
1250
Figure 1(a) Original ECG Signal
0 200 400 600 800 1000 1200
900
950
1000
1050
1100
1150
1200
Figure 1(b) Reconstructed ECG Signal
In a similar way, different samples of ECG signal are taken
from MIT- BIH arrhythmia data base (record 100) and for a
particular threshold value (λ), the Compression Ratio (CR)
and the error parameter i.e. Percentage Root Mean Deviation
(PRD) is calculated.
Table 1 shows the different ECG samples of record 100 and
the calculated values of compression ratio and percentage root
mean deviation.
Table 1 Different ECG samples and the calculated parameters
Number
of
samples
Discrete Wavelet Transform
( Biorthogonal 4.4 wavelet – level 5)
Threshold
Compression
Ratio (CR)
Percentage
Root Mean
Deviation
(PRD in %)
1024 280.7 34 2.37
2048 251.3 28.05 2.02
4096 252 30.1 2.068
8192 239.8 31.03 2.047
16384 267 31.5 2.25
Figure 2 (a) shows the original ECG signal and (b) shows the
reconstructed signal for 16384 samples of record 100 from
MIT-BIH arrhythmia database with Threshold λ=267,
Compression Ratio = 31.5, Percentage Root Mean Deviation =
2.25%

__________________________________________________________________________________________
0 2000 4000 6000 8000 10000 12000 14000 16000 18000
850
900
950
1000
1050
1100
1150
1200
1250
Figure 2(a) Original ECG Signal
0 2000 4000 6000 8000 10000 12000 14000 16000 18000
850
900
950
1000
1050
1100
1150
1200
Figure 2(b) Reconstructed ECG Signal
CONCLUSIONS
From the results it can be concluded that as threshold value (λ)
increases, the Compression Ratio (CR) increases and the error
criterion Percentage Root Mean Deviation (PRD) value also
increases which yields high data reduction and poor signal
fidelity. Low threshold value gives low data reduction and
high signal fidelity. So threshold value should be selected such
that the quality of the ECG signal is not distorted on
reconstruction and a good amount of data reduction is also
achieved. The same data compression algorithm is to be
implemented in FPGA using Verilog HDL.
REFERENCES:
[1] Mrs.S.O.Rajankar and Dr. S.N. Talbar, ―An Optimized
Transform for ECG Signal Compression‖, ACEEE Int. J. on
Signal & Image Processing, Vol. 01, No. 03, Dec 2010
[2] Ruqaiya Khanam and Syed Naseem Ahmad,‖ ECG Signal
Compression for Diverse Transforms‖, ISSN 2224-5758
(Paper) ISSN 2224-896X (Online, Vol 2, No.5, 2012
[3] Prof. Mohammed Abo-Zahhad,‖ ECG Signal Compression
Using Discrete Wavelet Transform‖, Vice-Dean for Graduate
Studies, Faculty of Engineering, University of Assiut, Egypt
[4] Hari Mohan Rai and Anurag Trivedi, ―De-noising of ECG
Waveforms based on Multi-resolution Wavelet Transform‖,
International Journal of Computer Applications (0975 – 8887)
Volume 45– No.18 May 2012
[5] Chengjun Zhang and Chunyan Wang, ―A Pipeline VLSI
Architecture for High-Speed Computation of the 1-D Discrete
Wavelet Transform‖, IEEE TRANSACTIONS ON
CIRCUITS AND SYSTEMS—I: REGULAR PAPERS,
VOL. 57, NO. 10, OCTOBER 2010
[6] M. JANARDAN and Dr. K ASHOK BABU, ―AN
EFFICIENT ARCHITECTURE FOR 3-D LIFTING-BASED
DISCRETE WAVELET TRANSFORM‖, M Janardan et al,
Int. J. Comp. Tech. Appl., Vol 2 (5), 1439-1458
[7] Mohammed Abo-Zahhad, SabahM. Ahmed, and Ahmed
Zakaria, ―An Efficient Technique for Compressing ECG
Signals Using QRS Detection, Estimation, and 2D DWT
Coefficients Thresholding‖, Hindawi Publishing Corporation
Modelling and Simulation in Engineering,Volume 2012,
Article ID 742786, 10 pages doi:10.1155/2012/742786
[8] Lecture Notes: Ali Shoeb and Gari Clifford, ―Chapter 16 -
Wavelets; Multiscale Activity in Physiological Signals‖,
HST582J/6.555J/16.456J Biomedical Signal and Image
Processing Spring 2005
[9] Apoorv Gautam1 and Maninder Kaur, ―ECG Analysis
using Continuous Wavelet Transform (CWT)‖, IOSR Journal
of Engineering Apr. 2012, Vol. 2(4) pp: 632-635
[10] MIT- BIH arrhythmia data base www.physionet.org
[11] www.mathworks.com
[12] www.intechopen.com
BIOGRAPHIES:
Karishma Qureshi received the B. E.
degree in Electronics and
Communication in 2003. She is
working as an Engineer-SD at Institute
for Plasma Research since 2005. She is
pursuing her Masters in Engineering
from L. D. College of Engineering in
Applied Instrumentation Department, Ahemdabad
Vinod Patel received M. Tech degree
from Indian Institute of Technology,
Roorkee in the year 2007. Currently he
is working as an Associate Professor and
Head in Applied Instrumentation
Department at L. D. College of
Engineering, Ahemdabad

Efficient data compression of ecg signal using discrete wavelet transform

More Related Content

What's hot (20)

Viewers also liked (15)

Similar to Efficient data compression of ecg signal using discrete wavelet transform (20)

More from eSAT Publishing House (20)

Recently uploaded (20)

Efficient data compression of ecg signal using discrete wavelet transform