Study and Investigate life threatening ECG arrhythmias using morphological patterns and Wavelet transform method

The work in this paper is to investigate the detection of another group of arrhythmias, which might not be need immediate attention but critically life threatening that decline health of heart and become cause of cardiovascular disease. The Electrocardiogram plays an imperative role to diagnose such cardiac disease. It is used to monitor and analyze all cardiovascular functioning and cardiac signal processing. But recorded ECG often contaminated by various noise and artifacts like power line interference, baseline wander and movement of patient muscles. Hence, for accurate diagnosis of heart disease and characterize normal rhythm from arrhythmic wave, the first step is to obtain clear ECG signal. As compared to conventional filtering methods, Wavelet transform has better denoising capability of such non-stationary signals, like ECG. Therefore, the proposed method used a discrete wavelet transform (DWT) based de-noising procedure for pre-processing of signal to obtain dynamic features and morphological patterns of ECG arrhythmia. The algorithm was implemented on MATLAB and used MIT BIH arrhythmia database.