基于运动单元动作电位序列的肌电解码与假肢连续运动控制

项目拟通过研究基于高密度表面肌电测量技术的MUAPt解码方法，探索基于运动单元动作电位序列的新型灵巧假肢生机接口及其集成方法：从理论分析和实验验证两方面入手，通过侵入式肌内肌电信号和高密度表面肌电信号的同步测量，研究基于表面肌电信号的MUAPt的反解及其解剖学位置标定方法，构建新一代的神经控制模型。分析人手运动参数与MUAPt的相关性，建立运动模式在MUAPt中的表征模型、多自由度运动的同步比例预测模型，使生机接口功能由离散模式控制转向连续运动控制。研究面向MUAPt反解的高密度表面肌电通道约减方法，建立多自由度实时灵巧假肢控制生机接口，实现人手腕部及局部精细自由度的连续运动控制，为高传输率生机接口及灵巧假肢研制积累科学基础。

This project proposes to investigate the technology of decoding the motor unit action potential train (MUAPt) from high-density surface electromyography (EMG). This technique will be applied for a novel dexterous human-machine interface and the corresponding system integration. High-density EMG signals and intramuscular EMG signals will be recorded concurrently to verify the decomposition algorithm and to extract spatial information of MAUPts. The relationship between hand movements and MUAPt will be analyzed to build a novel control model which can realize simultaneous and proportional estimation of multiple degrees-of-freedom (DOF) kinematics. The recording channel configuration of high-density surface EMG, including electrode number and spatial distribution, will be optimized to improve the decomposition performance of MUAPt. This project will not only realize the simultaneous and continuous control of prosthetic hand wrist and fingers, but also provide enabling techniques for human-machine interface and dexterous prosthetics with high transmission rate.