Assessing Forearm Exertion in Manual Tasks with Surface EMG: A Comparative Analysis of Through-Forearm vs. Muscle Specific EMG Placements

Abstract

BackgroundHand-intensive work is associated with musculoskeletal disorders (MSDs) such as carpal tunnel syndrome, tendinitis, and nerve entrapment. Estimating internal load on tissues like tendons and muscles in the hand and forearm is essential but challenging. Consequently, external force or self-reported exertion are often used as proxies for internal load assessment. Surface electromyography (sEMG) provides a promising method for continuous and objective monitoring of force. Through-forearm sEMG - placing one electrode on the flexors and one on the extensors - has been suggested as a pragmatic approach for capturing physical load under different wrist and forearm postures. However, further validation is needed to determine its accuracy in predicting exertion and force across diverse tasks, as well as to understand what specific aspects of muscle activity it captures. This study aimed to compare through-forearm sEMG with two common muscle-specific placements, i.e., the forearm extensor and flexor, in estimating force and perceived exertion during hand-intensive tasks and to explore the correlations of sEMG signals between these three sEMG placements. MethodsSixteen participants performed four simulated tasks, i.e. gripping, thumb-pressing, screwing, and key-pinching, at five exertion levels following a modified OMNI-Resistance Exercise Scale (OMNI-RES). The sEMG signals from the through-forearm, extensor, and flexor placements were recorded alongside exerted force. Polynomial mixed-effects models were used to estimate self-rated exertion and exerted force from muscle activity. Pearson correlation and linear mixed-effects models were used to explore correlations of sEMG signals between these three sEMG placements. ResultsAll sEMG placements predicted self-rated exertion and exerted force with strong model fit (R2 > 0.95 in most tasks) and high precision (residual SD < 1 for self-rated exertion, and < 5% for force in most tasks). However, moderate inter-subject and inter-task variability was observed. Through-forearm sEMG consistently outperformed both extensor and flexor placements across tasks by a small margin. The sEMG signals obtained from the three sEMG placements were highly correlated, and through-forearm sEMG could be expressed as a linear combination of extensor and flexor signals. DiscussionThe findings suggest that exertion estimation with a single-channel sEMG is feasible, although individual calibrations and additional task-specific information may be necessary for real-world applications. Through-forearm sEMG placement marginally outperformed extensor and flexor placements, which could be due to its integration of muscle activities from both muscle groups. ConclusionThe through-forearm sEMG placement provides slightly better performance estimating exertion than muscle-specific placements in the assessed manual tasks, possibly due to its integration of information from multiple muscles. Further research is needed to explore methods for individual calibration and ways to obtain task-specific information to enhance applicability in broader real-world settings.