Abstract
Correct-active-repetitive exercise is important for the motor recovery in hemiplegics. The present study hypothesizes that the electromyogram (EMG)-controlled assistance of motion would be an effective implementation of the concept for the rehabilitation of the hemiplegic shoulder, even in chronic patients. This study aims to investigate the feasibility of the suggested method. The motor intention is derived from the EMG of the shoulder muscles and the shoulder movement (flexion and abduction) is assisted by an electro-mechanical system only when the motor intention (EMG amplitude) exceeded the threshold. Twelve patients in the chronic stage of stroke participated in this pilot study. The EMG-controlled assistive exercise lasts for two weeks, 20 min per day and 5 days a week. The active range of motion in both abduction and flexion increases significantly after the intervention (p < 0.01). The maximum torque increases in both directions, and the increase is significant in the abduction (p < 0.01). The Fugl–Myer motor assessment score is improved greatly in the shoulder-related items (p < 0.01), but neither in the shoulder-unrelated items of the upper extremity (p = 0.13) nor in the lower extremity items (p = 0.19). This pilot study demonstrates that EMG-controlled assistive exercise can improve shoulder motor functions related to selected muscles and the suggested method is promising for the motor recovery of the shoulder in chronic hemiplegia.