EMG pattern classification for a prosthetic forearm with three degrees of freedom

Abstract

The authors developed an EMG controlled prosthetic forearm with three degrees of freedom actuated by small size ultrasonic motors. Its weight is less than 700 g and the size is the same as the adult's forearm. The interface between the amputee and the prosthetic forearm was designed on the basis of the fact that the amputee still preserves the phantom limb motor map after amputation. The paper proposes a method to estimate the motion intended by an amputee from the EMG signals using neural network. The method presented can discriminate the amputee's intended motion among six kinds of limb-functions from the multi-channel EMG signals preprocessed by the bandpass and smoothing filters. The cross-information among the EMG signals can be utilized to make the electrode locations flexible, and the band-pass filters can provide the amplitude and frequency characteristics of the EMG signals. The experiments of three subjects and four electrode locations demonstrates that the method can discriminate six motions of forearm and hand from unlearned EMG signals with the accuracy above 90 %. >