A kinematic-muscle hybrid synergy-assisted EMG-driven NMSK model for muscle activation and joint moment prediction with minimal sensors
Biomedical Signal Processing and Control · 2027-02 · 130 · 111575
DOI: 10.1016/j.bspc.2026.111575 ↗Kurz gefasst
Fachartikel in der Zeitschrift Biomedical Signal Processing and Control, 2027-02. Verfasst von Xue Chen und 1 weiteren Autor:innen. Die Arbeit nennt 38 Literaturangaben. Gefördert von: National Key Research and Development Program of China.
Abstract (Originalsprache)
Cross-country sit-skiing relies heavily on upper-limb power, with double poling (DP) propulsion generated by coordinated pole forces and trunk rotation. Estimating shoulder and elbow joint moments and muscle forces during DP is crucial for technique optimization and injury-risk reduction; however, electromyography (EMG)-driven neuromusculoskeletal (NMSK) models typically require many EMG channels, limiting feasibility in real-world competitions. To address this limitation, this study developed a kinematic-muscle hybrid synergy-assisted EMG-driven NMSK (KMHS-EDN) model that predicts muscle activations and joint moments using EMG recorded from a minimal set of muscles. We designed 29 input muscle combinations comprising 3–5 muscles each. Using calibration data, the KMHS-EDN model was calibrated for the motor module, musculotendon, and activation-dynamics parameters by fitting experimental joint moments, and was then used to estimate unmeasured muscle activations. Performance was evaluated using uncertainty quantification (confidence intervals for error metrics) and effect size-based quantification of prediction accuracy based on error magnitudes. Results indicated that the proposed KMHS-EDN model provided broadly comparable predictive quality when driven by EMG signals from infraspinatus, posterior deltoid, triceps brachii, and teres major, particularly for reproducing unmeasured muscle activations and shoulder-elbow joint moments during DP propulsion. Within-subject evaluation further supported model stability and identified this four-muscle subset as the best-performing minimal combination among those tested. By reducing the number of required EMG channels from 10 to 4 (60 % reduction) without a disproportionate loss in predictive performance, this preliminary framework may simplify experimental measurement and improve computational efficiency, supporting practical applications in sports biomechanics and clinical rehabilitation.
Förderung
- National Key Research and Development Program of China
Zitieren
APA
Chen, X., & Huo, B. (2027). A kinematic-muscle hybrid synergy-assisted EMG-driven NMSK model for muscle activation and joint moment prediction with minimal sensors. Biomedical Signal Processing and Control, 130, 111575. https://doi.org/10.1016/j.bspc.2026.111575
Vancouver
Chen X, Huo B. A kinematic-muscle hybrid synergy-assisted EMG-driven NMSK model for muscle activation and joint moment prediction with minimal sensors. Biomedical Signal Processing and Control. 2027;130:111575. doi:10.1016/j.bspc.2026.111575
BibTeX
@article{Chen2027,
author = {Chen, Xue and Huo, Bo},
title = {A kinematic-muscle hybrid synergy-assisted EMG-driven NMSK model for muscle activation and joint moment prediction with minimal sensors},
journal = {Biomedical Signal Processing and Control},
year = {2027},
volume = {130},
pages = {111575},
doi = {10.1016/j.bspc.2026.111575}
}
EUVIMED