Forearm Movements Classification Research to Increase Subjects Independence

Abstract

The difficulty of current pattern recognition is its applicability. Many manuscripts with high recognition accuracy are based on extensive training in the laboratory. In practice, it is impossible to carry out a large amount of exercise for every participant, so how to reduce the dependence on participants has become the focus of current research. This study introduced a Fuzzy C-Means (FCM) algorithm to realize the forearm movements' recognition to increase subjects' in-dependence. The method could be used between individuals; that is, every participant could select the actions by himself, breaking the traditional defect that could only identify the specific activities. The research paper shows effective methods from 2-channel electrodes data collected and analyzed from 8 participants. These participants selected five movements, and the average accuracy was 80.26%. It suggests that the control strategy chosen could be employed on different individuals. This method can promote the development of rehabilitation training for patients with muscle weakness.