Trans Humeral Prosthesis Based on sEMG and SSVEP-EEG Signals*

Abstract

The loss of forearm muscle in amputees above elbow joint make it impossible to control the prosthesis of elbow joint and upper limb only by using surface electromyography (sEMG) signals. Electroencephalogram (EEG) signals can be used as input signal to control the motion of the upper limb prosthetic hand for it can reflect the user's motion intention. This paper introduces a method of controlling the trans humeral prosthesis by combining sEMG and EEG signals. In this method, the control of elbow flexion and extension motions are based on sEMG signals of biceps and triceps. Combined with the collected elbow angles, the elbow angle of prosthetic arm is predicted by back propagation neural network after training and then the angle can be used to control the elbow joint. In order to control the motion of the prosthetic hand, a control method based on EEG is proposed. The EEG control method is named as steady state visual evoked potential (SSVEP). User can use his EEG signals to control the motion of hand prosthesis. Canonical correlation analysis (CCA) algorithm is used to classify SSVEP signals, then different SSVEP signals can be used to control different motions of prosthetic hands. Some experiments were carried out on healthy subjects to verify the performance of the proposed system.