WAVELET SHRINKAGE-BASED COEFFICIENT FOR ARRHYTHMIA DETECTION IN ECG DATA

Abstract

This research explores the application of wavelet shrinkage for denoising electrocardiography (ECG) data and detecting arrhythmias. Wavelet shrinkage is a nonparametric regression technique used to reduce noise in signals by performing wavelet transformations and attenuating high-frequency components. The focus of this study is on RR time series, which represents the interval between successive R-peaks in ECG data. We introduce a novel approach called Coefficient of Arrhythmias Detection based on Wavelet Shrinkage (CADWS), incorporating the coefficient of variation for enhanced accuracy. To evaluate CADWS, we compare 16 healthy and 16 unhealthy RR time series from the MIT-BIH database. Our findings suggest that CADWS can serve as a valuable diagnostic tool for identifying cardiac arrhythmias, contributing to improved healthcare diagnostics.