Detecting the motor unit action potential from surface EMG signals based on wavelet transform

Due to the poor temporal resolution of surface EMG signals, it is a challenging task to detect the motor unit action potential (MUAP) directly from the surface EMG signals. In this paper, a novel method based on wavelet transform was proposed to enhance the temporal resolution. In the method, some thresholds were applied to process all of the wavelet coefficients at every scale, and then the signals of every scale were respectively reconstructed with their own processed coefficients. Instead of wavelets, some isolated waves were chosen from these reconstructed signals as their templates, and then author respectively calculated the convolution of these templates with their own reconstructed signals. As a result, the firing pattern (FP) of the MUAP or multiunit burst could be determined from the convolution with a pre-set threshold.