上肢可穿戴式同源自主肌电控制功能性电刺激康复系统关键技术研究

Voluntary-controlled functional electrical stimulation (FES) can improve the motor function of the upper extremity post-stroke significantly. The FES controlled by the electromyographic signals (EMG) from the homologous muscle is such an effective way to enhance voluntary control. In contrast to natural force exertion that EMG amplitude and the muscle force are in a quasi-linear relationship, the muscle force generated by homologous EMG-controlled FES are the non-linear summation of the voluntary force and the stimulus-induced force. Therefore, the present EMG-controlled FES methods, which only control the stimulation intensity linearly according to the EMG amplitude envelope, is not able to control the muscle force naturally. Recent studies found that the muscle force can be mimicked with high fidelity by co-modulating stimulation intensity and frequency simultaneously and the stimulus-induced muscle fatigue can be alleviated. However, because the existing methods of stimulation noise cancellation are not applicable for the dynamic frequency stimulation, the voluntary EMG cannot be extracted during the co-modulation of stimulation intensity and frequency. In order to solve the above problems, firstly, a stimulation noise cancellation method for the stimulations with dynamic parameters will be designed in this project. Secondly, the relationship between the EMG feature，the stimulation parameters and the hybrid muscle force will be studied, so as to build a mathematical model for the linear control of hybrid force according to the voluntary EMG. Finally, a wearable EMG-controlled FES system will be designed scientifically to help the stroke patients controlling the muscle forces naturally. Thus, they can better promote the recovery of upper extremity through target-oriented rehabilitation tasks.

自主意识控制的功能性电刺激(FES)可显著地提高脑卒中的上肢康复疗效，同源肌肉自主肌电(EMG)控制正是自主意识控制FES的有效途径之一。由于自然发力时的EMG幅值和肌肉力量呈近线性的关系，而同源EMG控制电刺激产生的力量为自主发力和刺激诱发力量的非线性叠加，现有的控制方法仅通过EMG幅值包络线性调控刺激强度是无法自然地控制肌肉力量的。近期研究表明，通过EMG特征同时调控刺激强度和频率是准确控制肌肉力量的有效途径，并可降低肌肉疲劳，但是现有的刺激噪声消除方法不适用于变频刺激产生的噪声，因此无法实现强度、频率双重调控下的自主EMG提取。本项目针对上述问题，首先提出动态参数刺激的噪声消除方法。在此基础上研究自主EMG特征、刺激参数和混合肌肉力量的关系，并建立自主EMG特征线性控制混合力量的模型。通过穿戴式系统的科学设计，使脑卒中患者达到类似自然的肢体控制体验，通过任务指向性训练提高康疗效。

自主意识控制的功能性电刺激(FES)可显著地提高脑卒中的上肢康复疗效，自主肌电(EMG)控制正是自主意识控制FES的有效途径之一。由于自然发力时的EMG幅值和肌肉力量呈近线性的关系，而同源EMG控制电刺激产生的力量为自主发力和刺激诱发力量的非线性叠加，现有的控制方法仅通过EMG幅值包络线性调控刺激强度是无法自然地控制肌肉力量的。近期研究表明，通过EMG特征同时调控刺激强度和频率是准确控制肌肉力量的有效途径，并可降低肌肉疲劳，但是现有的刺激噪声消除方法不适用于变频刺激产生的噪声，因此无法实现强度、频率双重调控下的自主EMG提取。本项目提出了两种适用于变频、变幅电刺激的动态刺激噪声去除方法， 设计了一种去除刺激伪迹的肌电探测前端电路，并使用嵌入式系统实现了动态参数刺激下实时肌电信号的提取。两种方法的综合性能，均超过现阶段已报道的刺激噪声消除方法。在此基础上，成功实现了闭环穿戴式多动作自主肌电控制FES系统，并完成了健康人和脑卒中患者自主肌电控制FES测试实验。此外考虑到商用肌电电极存在力学失配、穿戴舒适度差等问题，基于本项目设计研发了一种基于Au@AgNWs/PVDF 超薄共型生物电信号探测体表电极，为将来的穿戴式系统的舒适性和长期稳定性打下了基础。