Intelligent myoelectric control of a humeral ampulation

Abstract

This paper presents the development of an intelligent system capable of managing the mobility of a myoelectric prosthesis for above elbow amputees. A choice of an acquisition protocol allowed us to produce two EMG signal databanks for two types of movement extension and flexion. Furthermore, the databanks were used to carry out the training of the intelligent system. With only two temporal characteristics, the SVM classifier test produced an accuracy of 84.75%. The performance of the latter has been validated using cross-validation. These results are promising in terms of real-time implementation of an intelligent embedded system of a myoelectric elbow prosthesis.