Development of an EMG controlled hand exoskeleton for post-stroke rehabilitation

Abstract

Rehabilitation after stroke is of crucial importance since patients often have severe motor impairments that affect their daily activities. In the past decades, a number of robotic systems have been proposed for stroke rehabilitation but very few of them are available in the market. Additionally, because of their high costs they are not widely accessible. This paper presents a light-weight hand exoskeleton controlled by patient's surface electromyography (sEMG) signals. This system could empower the patient to intensively exercise the hand and control its movement with his/her muscles. This would personalise the rehabilitation process since the system will respond to patient's own effort to move. sEMG signals were acquired from the healthy forearm and pattern recognition techniques were used to distinguish extension and flexion of the index and middle fingers. The motorized exoskeleton imposed the corresponding movements in the hand wearing it. Individualization of finger movements was accomplished with more than 95% of accuracy and opening and closing motions were established by the exoskeleton in the opposite hand.