Continuous wavelet transform application to EMG signals during human gait

Abstract

EMG signals are important in quantifying deviations from normal gait. Traditionally, Fourier transforms were utilized in determining the frequency spectrum of the typically non-stationary EMG signals. The continuous wavelet transform, suggested in this paper, is more appropriate. In this study, signals from four muscles of the right lower extremity were recorded, for eight normal subjects, during steady-state gait. The time-frequency distributions of these signals were computed using the fourth order Daubechies mother wavelet. Wavelet-based time-frequency representations were useful in identifying the recruitment patterns of slow and fast fibers to meet the varying demands imposed on the muscles during different phases of the gait cycle.