Non-invasive estimation of the degree of motor unit synchronization in the biceps brachii muscle

Abstract

This paper addresses estimation of motor unit (MU) synchronization by means of surface electromyogram decomposition. Firstly, the so called convolution kernel compensation method for identifying the discharge patterns of individual motor units is briefly described. The method builds on independent component analysis framework and is, hence, highly robust to superimpositions of MU action potentials (AP). In this study, the method was tested with synthetic signals in presence of MU synchronization, when the MU discharges are not strictly independent anymore. The level of MU synchronization was measured by the cross-interval histograms of reconstructed MU discharge patterns. Two synchronization indices were computed: the average number of synchronized MU pairs per contraction and the average percentage of synchronized AP per synchronized MU pair. Results revealed that both indices vary substantially with the properties of reconstructed MUs, but can nevertheless be used as reliable estimators of MU synchronization. The method was then applied to experimental electromyographic signals, acquired during low force contractions of the dominant biceps brachii muscle. Up to 14 concurrently active MUs were identified. MU synchronization was observed in approx. 50% of MU pairs. It is thus concluded, that MU synchronization can be assessed by decomposing the surface electromyogram.