Subject-specific EMG pattern classification ofactive hand movements for prosthesis applications

Abstract

The prosthesis hand amputees are highly helpful for various active hand movements based on wrist and elbow mobility for specific subject. In the field of rehabilitation, development of an advanced human-machine interface has been an interesting research topic in which biomedical electromyography (EMG) signals, play a vital role. Capturing, pre-processing, feature extraction and classification of EMG is very desirable which allows more standardized and precise evaluation of the neurophysiological, rehabitational and assistive technological advancements in prosthetic applications. This paper concerns with the capturing of real-time active hand movements EMG signals based on wrist-elbow mobility for simultaneous classification of features. The Anterior and Posterior forearm muscles are considered for proficient manipulation of EMG signals. The Feature is extracted using statistical first order time-frequency scaling analysis with pattern classification via linear discriminant analysis (LDA) which estimates the classification rate of about (89-91)%.