Motor Unit Discharges from Multi-Kernel Deconvolution of Single Channel Surface Electromyogram

Surface electromyogram (EMG) finds many applications in the non-invasive characterization of muscles. Extracting information on control motor units (MU) is difficult when using single channels, e.g., due to low selectivity and large phase cancellations MU action potentials (MUAPs). In this paper, we propose a new method face problem case differential channel. The signal approximated as sum convolutions different kernels (adapted signal) firing patterns, whose estimation cumulative firings. Three simulators were used for testing: muscles parallel fibres with either two innervation zones (IZs, thus, MUAPs phases) or one IZ model inclined respect skin. Simulations prepared fat thicknesses, distributions conduction velocity, maximal rates, synchronizations discharges, variability inter-spike interval. performances measured terms cross-correlations estimated simulated firings range 0–50 Hz compared those state-of-the-art single-kernel algorithm. median multi-kernel/single-kernel approaches 92.2%/82.4%, 98.1%/97.6%, 95.0%/91.0% models IZs, (parallel fibres), fibres, respectively (all statistically significant differences, which larger MUAP shapes greater difference).