An exoskeleton upper limb rehabilitation robot based on electroencephalography

Abstract

Today, stroke patients have varying degrees of motor impairment after surgery. An Electroencephalography (EEG) signal is a potential change recorded on the scalp of a human or animal, which can be combined with a rehabilitation robot to help patients complete rehabilitation movements. In this paper, a new exoskeleton-type 6-DOF upper limb rehabilitation robot is designed based on EEG control. The wavelet denoising method based on Gaussian mixture model (GMM) is used for signal pre-processing. The wavelet packet decomposition method is used to extract feature vectors, and the feature performance index based on Mahalanobis distance and Babanobis distance is introduced to test the accuracy of Feature Performance Index (FPI) relativity. The random forest classifier was used to classify and recognize the EEG characteristics and obtain the motion intention of patients. The experimental research shows that the EEG signal processing method proposed in this paper has significant effect, and the upper limb rehabilitation robot based on EEG signal has feasibility. The whole system can significantly improve the patient’s rehabilitation enthusiasm.