Study on Best Wavelet Packet Based Independent Threshold De-noising for MUAP

Abstract

Currently, wavelet packet technology is widely used in Electromyography (EMG) signal de-noising. However, most of these methods use global threshold to remove the noises from EMG signal. This paper proposes a new method, different nodes use different thresholds. The detected EMG signal is the summation of motor unit action potential (MUAP) trains from all active motor units. In our method, we use emglab software to abstract the MUAP of EMG, then calculate the best wavelet package tree and deal with each terminal node by independent threshold. Finally, the signal can be reconstructed by these processed coefficients. Compared with the wavelet packet de-noise with global default threshold, wavelet packet de-noise with level independent default threshold, the proposed method has distinguish advantageous.