An efficient wavelet based ECG de-noising using level dependent thresholding

This research is performed for the elimination of various noises from ECG signal. In this research work an efficient wavelet has been proposed. The coefficients of the proposed wavelet have been tuned to maximize the denoising of ECG signal. Soft thresholding has been used as it gives high accurate results rather than hard thresholding. The results have been obtained on self-acquired ECG data set and MIT standard data bases. This method was compared with the previous methods like Donohos universal method, s-median thresholding, modified s-median thresholding, etc. The comparisons lead us to the conclusion that our proposed method gives much better results in terms of various analysis parameters like MSE, SNR-improvement, etc.