Wavelet Based ECG Signal De- Noising

Abstract--

ECG is a method of measuring, displaying and recording the electrical

activity of heart. Wavelet de-noising method eliminates noise from the ECG signal. Using
soft and hard thresholding process, the significant wavelet coefficients from each level
are selected and denoised. ECG signal is reconstructed with inverse wavelet transform.
Different thresholding algorithms are analyzed and their performances have been
evaluated in terms of SNR. The standard MITBIH arrhythmia data from Physionet is
used for the purpose.

Introduction:
The electrocardiogram (ECG) signal is the electrical interpretation of the heart activity; it
consists of a set of, well-defined, successive waves denoted: P, Q, R, S, and T waves. A
great intention has been paid to the adequate and accurate analysis of the ECG signal that
would lead to cardiac anomalies diagnosis. However, as the major part of real signals; the
real picked-up ECG signal is corrupted by several sources of noise: EMG
(electromyogram) signal (a high frequency signal related to muscle activity), the BLW
(the baseline wandering: a low frequency signal caused mainly by the breathing action),
the electrode motion (usually represented by a sharp variation of the baseline) ...This
corrupted noise prevents considerably the accurate analysis of the ECG signal and useful
information extraction. Different works have been established to design filtering
algorithms aimed to improve the SNR (signal to noise) values and recovering the ECG
waves in different noisy environment. Recently, the wavelet theory de-noising have been
widely exploited in noisy ECG filtering. Several wavelet de-noising ECG signal
algorithms were developed, exploring each a particular parameter: the wavelet function,
threshold calculus, and level decomposition. In this context, we develop in this present
work a de-noising wavelet algorithm based on the corrupted WGN (white Gaussian
noise) estimation.

Literature:
In ECG De-noising, it is clear that the precise detection of onset and offset of its main
waves in the noisy signal becomes very hard. Note that the sampling rate of the signals is
1 kHz (According to the PTB diagnosis ECG database for Physio)While there are welldeveloped methods for power line, interference, and drift suppression, there are still
problems in myopotential signal suppression due to the considerable overlapping of the
frequency spectra of both types of signals. Thus, the automatic interpretation, fol- lowing
accurate detection of characteristic ECG points and waves such as P wave, T wave, QRS
complex and the measurement of signal parameters become an extremely difficult,
sometimes virtually impossible task2.Generally, adequate ECG de-noising algorithms and
procedures should have the following properties1:a) Improve signal-to-noise ratio (SNR)
for obtaining clean and readily observable recordings, yielding the subsequent use of
straightforward approaches for ac- curate automatic detection of characteristic points in
the ECG signal and recognition of its specific waves and complexesb) Preserve the
original shape of the signal and especially the sharp Q, R, and S peaks, without distorting
the P and T waves and the smooth transition of the ST-T segment.
Recently some new techniques based on global and local transforms have become
popular in connection with signal denoising. At the first step, the signal is decom- posed
into a transform domain where filtering procedures are applied. The noise-free signal is
then obtained by an inverse transform. Choosing appropriate basis functions for
successful decorrelation of the signal and designing transform domain filters
accommodated to the ECG signal morphology could turn these techniques into powerful
means for ECG signal denoising.A method for ECG denoising has been proposed based
on the wavelet shrinkage approach3 using time-frequency dependent threshold (TFDT).
The TFDT is high for non-informative wavelet coefficients, and low for informa- tive
coefficients representing the important signal features. Although giving relatively good
results in comparison with other ECG denoising methods, the latter has some disadvantages: Some oscillations may occur at the end of the QRS-complexes using longlength decomposition filters due to the poor time localization of the basis functions; and
in the opposite direction very short-length filters may corrupt the shapes of the slow P
and T waves1.Using artificial neural networks (ANNs) for noise and baseline removal is
also used in many articles; e.g., a neual network based adaptive algorithm for ECG
denoising is presented. Furthermore, G.Clifford presented a general technique using
energy thresholding and Gaussian kernels for biomedical signal denoising and he showed
that significant noise reduction, compression, and turning point location is possible by
this method. Also, there are many mathematically based algorithms using wavelet
analysis to denoise ECGs and these can be found in many articles and papers.Despite
these mentioned algorithms, there are also problems in ECG denoising to preserve ECG
morphology with a high degree of confidence. In the present study, we aim to improve
the denoising procedure by using a mathematical algorithm.

Problem Definition:
ECG is a method of measuring, displaying and recording the electrical activity of heart.
This project aims to address the wavelet based de-noising technique for noise corrupted
ECG signals. The main focus of this project to estimate the noise in the given noise
corrupted ECG signal. With this estimated noise, the wavelet threshold is calculated and
this calculated threshold used for de-noising the wavelet coefficient in the given level.
Inverse wavelet transform is applied on de-noised wavelet coefficients to reconstruct the
de-noised ECG signal. This project also compares with various threshold computation
techniques like universal threshold, sure shrink, and Bayes shrink, and also thrsholding
techniques like with hard thresholding, soft thresholding, and hybrid thresholding.

Proposed Method:
In DWT the filters with different cutoff frequency are used for the analysis of the signal
at different scales. At each level of decomposition the high pass filter gives detailed
coefficients, which generally represent the noise of the signal and output of the low pass
filter gives approximation coefficients, also called coarse approximation. The threshold is
estimated based on the noise variance and signal variance in each level then soft and hard
thresholding is performed. Then the de-noised signal could be reconstructed from the
resulting wavelet coefficients and corresponding steps are depicted below.
1. Transform the noisy ECG signal into wavelet domain for finding DWT
coefficients in each decomposition level (Also called sub-band).
2. Estimate the noise inherent in the given noise corrupted ECG signal.
3. Computing the approximated threshold based on the noise variance and signal
variance in each sub-band.
4. Applying thresholding to obtain the estimated wavelet coefficients for each level.
5. Reconstruct the de-noised ECG signal .

ECGECG ++
ApplApplyy DWTDWT
noinoissee

Softwares Required:
Database: MITBIH

esesttiimmatatee
tthhee noinoissee

ccoomputmputee
tthhee
tthhrreesshhololdd

tthhrreesshhololddiinn
gg wiwitthh
ccoomputmputeedd
tthhrreesshhololdd

iinnververssee
eaveleaveleett
ttrraassffoorrmm

De-De-nnoioisseedd
ECGECG ssiiggnalnal

Matlab, MSVC++

Conclusion:
We analyze a very important signal, the electrocardiography by applying an advanced
filtering tool called discrete wavelet transform. ECG signal efficiently to keep it
distortion free and smooth. In signal denoising applications, soft thresholding method
gives better results than hard thresholding method. From simulation results we can
observe that the wavelet transform can remove the noise effectively and improve the
SNR. By using rigrsure threshold and minimaxi threshold we can effectively reduce the
noise and retain the useful information of the ECG signal. Use of heursure and sqtwlog
threshold techniques will remove a useful component of ECG signal. The advantage of

the wavelet based de-noising method is that it retains both the geometrical characteristics
of the original ECG signal and variations in the amplitudes of various ECG waveforms
effectively.

References:
1.
2.
3.
4.
5.

Ergun Ercelebi, "Electrocardiogram signals de-noising using lifting-based

discrete wavelet transform" Elsevier, Computers in Biology and Medicine vol. 34,
pp. 479-493, 2004.
Lukas Smital, Martin Vtek, Jir Kozumplk and Ivo Provamk, Member,
IEEE "Adaptive Wavelet Wiener Filtering of ECG Signals", IEEE Trans. on
Biomed. Engg., vol. 60, no. 2, pp. 437, Feb 2013.
Md. Ashfanoor Kabir, Celia Shahnaz, "Denoising of ECG signals based on noise
reduction algorithms in EMD and wavelet domains'~ Elsevier,Biomedical Signal
Processing and Control vol. 7, pp. 481-489, 2012.
M. S. Manikandan, S. Dandapat, "Wavelet energy based diagnostic distortion
measure for ECG", J. Biomed. Signal Process. Control,Elsevier 2 , pp. 80-96,
2007.
Jianbo Gao, Member, IEEE, Hussain Sultan, Jing Hu, Member, IEEE,and WenWen Tung, "Denoising Nonlinear Time Series by Adaptive Filtering and Wavelet
Shrinkage: A Comparison", IEEE Sign. Proce.Let., vol. 17, no. 3, pp. 237-240,
Mar. 2010.