Decomposition of EMG signal by wavelet spectrum matching

We have developed a technique using wavelet transform to classify single motor unit (SMU) potentials and to decompose EMG signals into their constituent SMU potentials. The distinction between this technique and previous techniques is that it measures waveform similarity of SMU potentials from wavelet domain which gives this technique many advantages over other techniques. Separation of compound motor unit potentials, on the other hand, is based on a sequence of waveform differential correlation processes. Our technique also utilizes the information on discharge regularities of SMUs. This technique had been tested using actual EMG recordings that were composed of more than 10 different motor unit potentials. The results were satisfactory