Denoising of ECG signals using weighted stationary wavelet total variation

•A total variation (TV) based ECG signal denoising is reviewed.•A new denoising method is proposed in SWT domain.•A new weighted denoising TV function is introduced.•The results and comparative analysis are discussed. ECG signals capture the electrical activity of the heart and can be used to determine its health. The P-QRS-T morphological features of the ECG may be used to identify a variety of cardiac problems. The existence of various disturbances in ECG signals, on the other hand, has a significant impact on correct feature extraction. During acquisition, various noises affect ECG readings, including power line interference, baseline wandering, motion artifacts, and electromyogram noise. In this paper, we proposed an algorithm called STationary Wavelet Total Variation (STWaTV) that introduces a total variation (TV) method with a bivariate shrinkage rule in the stationary wavelet transform (SWT) domain. STWaTV is tested with three existing threshold schemes: firm threshold, hard threshold, and bivariate shrinkage function with SWT for thresholding coefficients and TV for minimizing an objective function. To evaluate the effectiveness of STWaTV, we consider the standard databases and perform the comparative study along with existing state-of-the-art methods. In terms of qualitative analysis of STWaTV, the most significant signal to noise ratio (SNR) is in the range 13.23 to 53.61, root mean square error (RMSE) in the range 0.0006 to 1.1000, and the percentage root mean square difference (PRD) in the range 0.095 to 1.200. In sum, STWaTV outperform its counterparts and denoise ECG data without altering the amplitude of the original signals.