EMG Probability Density Function Inverse Problem: Motor Unit Potential Amplitude Versus Motor Unit Recruitment Threshold Maps.

Solving the EMG probability density function inverse problem is a novel approach to extract motor unit information from single-channel EMG recordings. The method operates on EMG signals recorded during an isometric ramp contraction covering the entire muscle force range. The EMG is segmented into small portions, and each segment is characterized using the first two non-central sample moments. An iterative approach is employed to successively identify newly recruited motor units in each segment. At each step, an optimization algorithm estimates the set of motor unit potential (MUP) amplitudes that provide an EMG probability density function model yielding the best fit to the experimental moments. The algorithm is evaluated using simulated signals, showing its ability to provide accurate estimations of MUP amplitudes versus recruitment thresholds. The method is also applied to real signals recorded from the tibialis anterior in 61 healthy and 71 pathological subjects. Comparison between healthy and pathological groups provides evidence of muscle fiber reinnervation and motor unit remodeling. The resulting MUP amplitude versus recruitment threshold maps constitute an innovative and informative tool to assess muscle state.