Multiscale wavelet energies and Relative Energy based Denoising of ECG signal

In this work, a novel denoising algorithm based on relative energies of Wavelet subbands and estimated noise variance is proposed for Electrocardiogram (ECG) signal. The proposed algorithm is based on Relative Energy Denoising (RED) factor which is a function of Energy Contribution Efficiency (ECE), Details Energy Contribution Efficiency (DECK) and the estimated noise variance of Wavelet subbands. The algorithm is tested with PTB Diagnostic ECG Database and CSE Multilead Measurement Library. Wavelet filtered signal fidelity is evaluated using Percentage Root Mean Square Difference (PRD), Wavelet Weighted Percentage Root Mean Square Difference (WWPRD) and Wavelet Energy Based Diagnostic Distortion (WEDD) Measure. The results are compared qualitatively and quantitatively with few existing gold standard thresholding methods. The proposed denoising method shows PRD, WWPRD and WEDD values as 5.0966, 14.8554 and 1.0677 respectively.