Mathematics of the muscle sound

Shows, in the case of the muscle sound, how an accurate analysis and selection of the mathematical methods adopted has led to a good description and interpretation of the physiological results. For all subjects, the power spectrum (PS) has been estimated both by FFT (fast Fourier transform) and MESE (maximum-entropy spectrum estimation). From the PS the mean, median, and peak frequencies have been computed. In all cases, it has been found that both the mean and the median frequency fill the requirement, the mean frequency being superior. A comparison of the three parameters at 20% and 80% maximum voluntary contraction (MVC) (no fatigue) is shown. The mean frequency of the power density spectrum seems to be adequate for the study of the sound myogram (SMG) in different physiological conditions, although it is not sensitive enough to allow monitoring of the recruitment of type-I (small, slowly fatiguing) and type-II (large, rapidly fatiguing) fibers. The cross spectrum of the electromyogram (EMG) and the SMG signal shows, by its similarity to the SMG power spectrum, that its information must be related to the activation pattern of the muscle fibers