Hurst´s rescaled-range (R/S) analysis and fractal dimension of electromyographic (EMG) signal

A microcomputer-based real-time signal acquisition system has been developed for online characterization of electromyographic (EMG) signals. A set of EMG signals is collected from three different patient groups: normal, neuropathic, and myopathic. The intramuscular signals are collected real-time for 2-3 s, during which the patient performs a continuous ramp contraction. The time-varying dynamic nature of the neuromuscular system is observed by fractal dimension measurement of the extended windowed data segments. The data are analyzed by Hurst´s rescaled-range-analysis method (H.E. Hurst, `Long-term storage: an experimental study´, Constable, London, 1965), and the Housdorff-Besicovich fractal dimension is calculated for each window length. The results of fractal dimension measurements for data from the different patient groups are then analyzed by using nonparametric statistical methods. A first-order regression model is used to quantify the trend of the model parameters. The Gaussian probability density functions are estimated from the empirical distribution of the model parameter, and the signals are classified on the basis of the probability density functions of the fractal dimension measurements