Adaptable EMG Prosthetic Hand using On-line Learning Method -Investigation of Mutual Adaptation between Human and Adaptable Machine

We developed a new adaptable EMG prosthetic hand, which executes recognition process and learning process in parallel and can keep up with change in the mapping between an electromyographic signals (EMG) to the desired motion, for amputee. EMG-to-motion classifier which used in proposed prosthetic hand is done under the assumptions that the input motions are continuous, and the teaching motions are ambiguous in nature, therefore, automatic addition, elimination and selection of learning data are possible. Using our proposed prosthetic hand system, we conducted experiments to discriminate eight forearm motions, with the results, a stable and highly effective discrimination rate was achieved and maintained even when changes occurred in the mapping. Moreover, we analyzed mutual adaptation between human and adaptable prosthetic hand using ability test and f-MRI, and clarified each adaptation process