基于表面肌电信号的上肢康复训练技术研究

Research on the Training Techniques of Upper Limb Rehabilitation Based on Surface Electromyography

为辅助上肢受损患者实现自主康复训练,促使肌肉信号与脑意识的再生通信,设计一种机械臂辅助式离散动作康复训练识别方法。受试者进行康复训练时,采集其肩关节4处肌肉群的表面肌电信号,提取时域特征,并采用BP神经网络分类算法对六种上肢肩关节动作意图进行模式识别。该方法能够准确建立表面肌电信号特征值与六种上肢康复动作之间关系映射模型,平均识别率高达90.27%。为基于表面肌电信号的外骨骼式自主康复训练系统提供一种可行的人机交互方案。

To help patients whose upper limbs are impaired realize independent rehabilitation training of upper limbs and make the regenerate communication between muscle signal and brain consciousness, a kind of auxiliary type mechanical arm rehabilitation training recognition method is designed for discrete movements. The surface electromyography （sEMG） of four muscles of the shoulder joint of the human body is been collected during the subjects＇ rehabilitation training. Then the time domain features are extracted, and the BP neural network classification algorithm implementation is used to recognize six kinds of shoulder joint action intention. The method can accurately establish the relational mapping model between the characteristic values and the sEMG from six kinds of shoulder joint motions, and the average recognition rate is as high as 90.27%. This scheme provides a feasible way of human-computer interaction for the exo- skeleton independent rehabilitation training system based on sEMG.