The probability density function of vibromyographic and electromyographic signals for different levels of contraction of human quadriceps muscles

Abstract

The probability density function (pdf) of random vibromyographic (VMG) and electromyographic (EMG) signals were studied. VMG and EMG signals were obtained from the quadriceps muscles of four subjects contracting at 20%, 40%, 60%, 80% and 100% of maximum voluntary contraction (MVC) at knee angles of 30/spl deg/, 60/spl deg/ and 90/spl deg/. Levels of contraction were controlled using the torque reading of a Cybex II dynamometer. Histograms of the VMG and EMG signals at different levels of contraction were made. Using Pearson's /spl chi//sup 2/-test, the goodness of fit to a Gaussian distribution was calculated for the VMG and EMG signals obtained from rectus femoris. The pdf of the VMG and EMG roughly approximated a Gaussian distribution, but both signals were found to be statistically different from a Gaussian distribution using Pearson's /spl chi//sup 2/ test. The goodness of fit of the pdf of the VMG and EMG signals to a Gaussian distribution were calculated. >