表面肌电信号控制的功能性电刺激康复系统研究

Study on Rehabilitation System of Functional Electric Stimulation of Hand Based on SEMG Signal Control

针对特定肌肉动作所需电刺激参数的差异以及组合动作中刺激位置顺序的排布,研发了一套以健侧肌电信号实时触发的多通道手部电刺激康复系统,以辅助脑卒中患者更有效的完成手功能康复训练。使用STM32F103微处理器完成下位机平台搭建,设计信号采集模块对表面肌电信号进行滤波及放大、利用表面肌电信号能反映肌肉运动信息的特点,使用Matlab获取手部运动特征建立分类模型、模型通过识别运动的健侧手部,将分类后的运动标签实时反馈至电刺激模块,以刺激患侧电极产生相应动作。最后经过重复性实验得到结果:系统的平均动作识别率可以达到90.7%;而特定手部肌肉动作的触发需要不同等级的电流强度。实验结果表明系统实时性良好、运行稳定、可实现模式识别后的相应刺激动作产生。

Aiming at the difference of electrical stimulation parameters needed for specific muscle movements and the arrange⁃ment of stimulation positions in combined movements,a multi-channel hand electrical stimulation rehabilitation system trig⁃gered by healthy side EMG signals in real time was developed where the output parameters of each channel can be adjusted inde⁃pendently to assist stroke patients to complete hand function rehabilitation training.The STM32F103 microprocessor is used to build the lower computer platform,and the hand movement characteristics are obtained by Matlab to establish a classification model where by identifying the moving healthy hand,the model feeds back the classified movement labels to the electrical stimula⁃tion module in real time.To stimulate the affected electrode to generate corresponding actions.Finally,the results show that the average action recognition rate of the system can reach 90.7%,while the trigger of specific hand muscle action needs different lev⁃els of current intensity.It shows that the system has good real-time performance and stable operation,and can generate corre⁃sponding stimulation actions after recognition.