Simplified EEWT Method for ECG Signal Analysis

This paper addresses mode decomposition for noisy ECG signals. We propose a simplified version of the Empirical Wavelet Transform (EWT) family of methods to select the useful spectral peaks. The simulations are carried out on actual electrocardiogram (ECG) recorded signals from standard Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH) arrhythmia database and normal sinus rhythm database. The proposed method is able to perform better mode decomposition with lesser computational complexity than the existing methods. Furthermore, R-peak detection is performed and useful parameters for different medical conditions are computed. We get accurate results even when the ECG signals are corrupted by baseline wander (BW) and power line interference (PLI).