Extraction of Motor Unit Dynamics using High-Density EMG for Neuromuscular Performance Quantification

The identification of skeletal muscle fatigue is imperative towards defining neuromuscular health and rehabilitation progress. While standard methods involving surface electromyography have proven capable of measuring fatigue at the muscle body level, understanding the behavior of individual motor units provides a deeper level of health information. By exploring motor unit activity, we determine the dynamic changes in recruitment and firing patterns over time. Moreover, we show that the decomposition of high-density electromyography signals can be used to analyze motor unit activity to identify muscle fatigue. When compared against a non-fatiguing task (i.e., minimal exertion) we show that the motor unit activity increases by 13.74\\% to 16.71\\% over a single trial. Across multiple trials, the average firing frequency decreases with more phasic behavior. Specifically, the motor unit dynamics become more non-linear, as the $R^2$ fit drops from an average of 0.56 for the non-fatiguing task to an average of 0.12 for fatiguing tasks. This research demonstrates a identification of neuromuscular fatigue at the motor unit level. Long-term applications of this research provide new tools and metrics in assessing rehabilitation, specifically when quantifying functional performance and longitudinal progression in neuromuscular disorders.