Muscle Activation–Deformation Correlation in Dynamic Arm Movements

Understanding the relationship between muscle activation and deformation is essential for analyzing arm movement dynamics in both daily activities and clinical settings. Accurate characterization of this relationship impacts rehabilitation strategies, prosthetic development, and athletic training by providing deeper insights into muscle functions. However, direct analysis of raw neuromuscular and biomechanical signals remains limited due to their complex interplay. Traditional research implicitly applied this relationship without exploring the intricacies of the muscle behavior. In contrast, in this study, we explored the relationship between neuromuscular and biomechanical signals via a motion classification task based on a proposed deep learning approach, which was designed to classify arm motions separately using muscle activation patterns from surface electromyography (sEMG) and muscle thickness deformation measured by A-mode ultrasound. The classification results were directly compared through the chi-square analysis. In our experiment, six participants performed a specified arm lifting motion, creating a general motion dataset for the study. Our findings investigated the correlation between muscle activation and deformation patterns, offering special insights into muscle contraction dynamics, and potentially enhancing applications in rehabilitation and prosthetics in the future.