Performance analysis of wavelet-based denoising techniques for ECG signal

Abstract

Scientific investigation concerning ECG (electrocardiogram) signal has attracted numerous researchers because of its pertinence for determining the condition of human's heart. Specific measurements using the ECG are commonly used by medical practitioners to portend early symptoms of heart disorders. Nevertheless, these measurements are often affected by unwanted noise which cannot be eliminated using simple filtering methods. Numerous studies have been conducted to develop ECG denoising techniques, yet there is no comprehensive investigation regarding their performance analysis. In this paper we provide a comprehensive simulation and analysis to measure the effectiveness of several wavelet-based denoising techniques. The simulation is performed using Matlab by assessing the values of Signal to Noise Ratio (SNR) and Mean Square Error (MSE). In our experiment, Adaptive White Gaussian Noise (AWGN) is added to the ECG signal prior to the denoising process. Our experiment examines three types of denoising techniques and yields the hard thresholding technique as the best method with MSE value of 0.000423 and SNR value of 28.8806 db.

9 Figures and Tables

Cite this paper

@article{Mandala2017PerformanceAO, title={Performance analysis of wavelet-based denoising techniques for ECG signal}, author={Satria Mandala and Yunendah Nur Fuadah and Muhammad Arzaki and Faida Esti Pambudi}, journal={2017 5th International Conference on Information and Communication Technology (ICoIC7)}, year={2017}, pages={1-6} }