Real-Time Prosthetic Hand Control Based on Muscle Synergy Decomposition of Forearm EMG Signals

Surface electromyography (sEMG) offers a noninvasive interface for prosthetic hand control, yet practical deployment remains limited by electrode shift, inter-subject variability, muscle fatigue, and repeated calibration requirements. This study proposes a real-time prosthetic control framework based on muscle synergy decomposition of forearm sEMG signals. Muscle activation levels were extracted from 16-channel sEMG recordings, and a fixed synergy matrix was obtained via non-negative matrix factorization (NMF) during an offline calibration phase. Real-time synergy activation coefficients were computed via pseudo-inverse estimation and mapped to control commands. A support vector machine (SVM) classifier was trained on time-domain features extracted from these coefficients. Experiments with six able-bodied subjects demonstrated that three muscle synergies consistently explained over 90% of data variance, with inter-subject cosine similarity exceeding 0.8. Offline classification achieved 93.03% mean accuracy across five gestures. In real-time object manipulation tasks, the system enabled successful power grasp, lateral pinch, and three-finger grasp without critical control failures. These results support the feasibility of synergy-based decomposition as a physiologically interpretable and computationally efficient solution for intuitive multi-functional prosthetic hand control.