On Detecting the Invariant Neural Drive to Muscles during Repeated Hand Motions: A Preliminary Study

A new method is proposed to detect the invariant neural drive from the central neural system (CNS) to forearm muscles during repeated hand motions. The method is based on a novel tracking approach for motor units decoded from high-density surface electromyography (HD-sEMG). Three healthy subjects performed eleven hand motions for eight trials. During the experiment, three high-density electrode grids were positioned around the forearm. The discharge timings of motor units are identified from HD-sEMG by a decomposition algorithm and used to trigger motor unit action potentials (MUAPs). We designed a tracking procedure and three different tracking sessions for MUAPs from forearm muscles, and we suggested a representative motor unit for each hand motion. Besides, motor unit pools are compared for different hand motions. Results show that motor units could be tracked during complex hand motions. The average peak tracking count across trials for the same motion is 98.1% higher than that for random sessions demonstrating the invariant neural drive between repeated hand motions. Besides, the average peak tracking count across motions in the same trial is 20.8% higher than random sessions, which means the same muscles are activated during different tasks. Representative motor units we proposed in this study have shown high discrimination between different hand motions. These results show the potential of building a more intuitive control device. We discuss the potential of learning the modular control strategy of the CNS during daily life hand movements and implement it to man-machine interfacing.