EMG model calibration and trajectory tracking by remote rehabilitation exoskeleton

This study presents a novel kinematic tracking model for an exoskeleton trained visually by therapist-guided exercises. Rehabilitation exercises are quantitatively assessed using a monocular passive vision system, and the exoskeleton uses analytical metrics from the camera as reference paths. Additionally, electromyography (EMG) feedback is incorporated, introducing a calibration approach to estimate angular positions based on EMG observations. The methodology is validated through vision-based metric data and experimental EMG measurements for exercises like shoulder flexion, elbow flexion, and rowing-like movements. Computer simulations demonstrate the system's adaptability in reliably, robustly, and effectively tracking rehabilitation exercise trajectories (Supporting video included).