Myopotential denoising of ECG signals using wavelet thresholding methods

We present empirical comparisons of several wavelet-denoising methods appli ed to the problem of removing (denoising) myopotential noise from the obser ved noisy ECG signal. Namely, we compare the denoising accuracy and robustn ess of several wavelet thresholding methods (VISU, SURE and soft thresholdi ng) and a new thresholding approach based on Vapnik-Chervonenkis (VC) learn ing theory. Our findings indicate that the VC-based wavelet approach is sup erior to the standard thresholding methods in that it achieves: Higher deno ising accuracy (in terms of both MSE measure and visual quality) and more r obust and compact representation of the denoised signal (i.e., it uses fewe r wavelets). (C) 2001 Elsevier Science Ltd. All rights reserved.