A combined muscle model and wavelet approach to interpreting the surface EMG signals from maximal dynamic knee extensions.

This study aimed to identify areas of reduced surface EMG amplitude and changed frequency across the phase space of a maximal dynamic knee extension task. The hypotheses were that (1) amplitude would be lower for eccentric contractions compared with concentric contractions and unaffected by fiber length and (2) mean frequency would also be lower for eccentric contractions and unaffected by fiber length. Joint torque and EMG signals from the vasti and rectus femoris were recorded for eight athletic subjects performing maximum knee extensions at 13 preset crank velocities spanning +/-300 degrees x s(-1). The instantaneous amplitude and mean frequency were calculated using the continuous wavelet transform time-frequency method, and the fiber dynamics were determined using a muscle model of the knee extensions. The results indicated that (1) only for the rectus femoris were amplitudes significantly lower for eccentric contractions (p= .019) and, for the vasti, amplitudes during eccentric contractions were less than maximal but this was also the case for concentric contractions due to a significant reduction in amplitude toward knee extension (p= .023), and (2) mean frequency increased significantly with decreasing fiber length for all knee extensors and contraction velocities (p= .029). Using time-frequency processing of the EMG signals and a muscle model allowed the simultaneous assessment of fiber length, velocity, and EMG.