Effects of SNR on removing ECG noise from EMG signal using DSWT

The electromyography (EMG) signal records the electrical potential difference across the skeletal muscle in every part of the body whereas the electrocardiography (ECG) signal measures the electrical potential difference across only the heart. For example, when the EMG signal is collected from muscles in the torso, the ECG signal coming up from the heart activity can interfere with the EMG signal. In this way, the EMG signal can be contaminated with the ECG interference during data collection. In this paper, we propose the improved optimal thresholding method based on discrete stationary wavelet transform (DSWT) in various 5-level SNR used for removing the ECG interference from the contaminated EMG signal. No prior studies have evaluated the performance evaluation of mean absolute error (MAE) and correlation coefficient (CC) for this proposed method to estimate the optimal threshold value in SNR 5 levels. The results show that the ECG interference is removed from the contaminated EMG signal by using the optimal wavelet thresholding method based on DSWT at various 5 levels of SNR. A proposed removal technique is better than the traditional thresholding method.