An improved denoise method of independent component analysis in the surface EMG signal

An improved denoise method based on independent component analysis (ICA) algorithm was discussed in this article to remove ECG noise in the surface EMG signal. In order to prevent loss of the useful EMG signal, a low pass filter was used to process the ECG components decomposed from the original surface EMG signal. The results of the simulation experiments suggest that the ECG components filtering with 30Hz low pass filter would be the most appropriate method to reduce possible surface EMG components before projecting them back. And then, projected the processed ECG components back to the original space, subtracted them from the original signal to get the EMG with no ECG. Experiment on the true 16-channel mono-polar surface EMG signal indicated that this method can achieve satisfying results in both time and frequency domain.