ロボット義手制御のための手指の筋シナジーに着目した欠損筋の活動推定

抄録

Various robotic prostheses have been developed to improve activities of daily living for amputees. Prostheses users can control fingers of the prostheses by using EMG of the extensor digitorum and flexor muscles remaining in the forearm. However, the interosseous muscle that performs adduction and abduction of the finger is located in the hand, so its EMG cannot be measured. The adduction/abduction motion of commercially available prosthesis is switched by attaching an additional switch. This does not allow the same versatile operation as the human hand. Therefore, in this study, we propose a method estimating the activity of the dorsal interosseous muscle of the index finger using synergy of finger muscles with an artificial neural network.