A virtual rehabilitation system based on EEG-EMG feedback control

Aiming at the limitations of the current rehabilitation training system in individual adaptability, safety and feasibility of operation, the electroencephalogram (EEG) and electromyography (EMG) signals which reflect the motion pattern and fatigue of patients are applied into the virtual rehabilitation system in this paper. The system collects the EEG and the surface electromyogram(sEMG) of upper limb in different gestures. The support vector machines (SVM) is applied to identify the action patterns after extracting the characteristics of the signals. Then the virtual scene and the target are controlled by the recognized motion intention. At the same time, the system achieved fatigue estimation based on EMG and EEG characteristics, and then adjust the difficulty coefficients of the virtual rehabilitation training scene to switch scenarios adaptively. The validity of the system is verified by experiments.