Motor unit recruitment during constant low force and long duration muscle contractions investigated with surface electromyography.

Surface EMG signals were detected from the biceps brachii muscle of five subjects using linear electrode arrays during isometric voluntary contractions at low force levels. A classification method based on neural networks has been used to identify the active motor units during the contraction. In addition, surface EMG global variables have been computed. It was found that at low contraction levels it is possible to reliably identify motor unit action potentials from the interference surface EMG signal and to classify them as belonging to different motor units. Progressive recruitment of new motor units during long duration contractions was observed in all the investigated cases from the first few minutes of contraction (3-4 minutes), indicating a change in the recruitment threshold of non-active motor units as a consequence of muscle fatigue. The recruitment of new motor units was more pronounced for higher force contraction levels than for the lower ones. This behaviour was confirmed in two out of five cases by changes of the global EMG variables.