A Method of Denoising ECG Signals Based on Improved Threshold and Modulus Maxima

This paper attempts to design a suitable wavelet denoising method for the clinic electrocardiogram (ECG) signals. Because ECG signal has better continuity and obvious singularity, the wavelet modulus maxima can be used to remove the EMG interference and respiratory wave from ECG signals. In order to get better denoising effect, the wavelet details (high frequency coefficients) are denoised by the wavelet threshold before doing modulus maxima analysis. Considering the advantages and disadvantages of hard threshold and soft threshold, an improved threshold function is designed and the appropriate denoising threshold is selected for each wavelet decomposing scale. Combining the wavelet improved threshold function denoising with the wavelet modulus maximum reconstruction, the effectiveness of this idea proposed by this paper is verified through MATLAB simulations, in the mean time, the denoising performance of the improved threshold function is verified to be better than hard threshold and soft threshold.