摘要
采用独立分量分析（ICA）技术和二阶非平稳源分离（SEONS）算法来研究肌肉轻度收缩情况下的表面肌电信号（SEMG）的运动单位动作电位（MUAP）检测问题，通过仿真实验来探讨两种算法对SEMG信号的分离性能，并将算法应用于肌肉轻度收缩时（10％MVC）的真实SEMG信号分解研究．仿真SEMG信号分解实验结果表明，两种算法对MUAP检测效果均较为满意，且随着噪声的增加有所变差，肌肉轻度收缩时（10％MVC）真实SEMG信号分解实验也论证了两种算法实际应用的可行性．盲源分离（BSS）技术为研究隐含在肌电信号中的运动单位募集和发放等信息提供了有效途径，较符合SEMG信号特性，因而可应用于SEMG信号分解及运动单位动作电位（MUAP）检测等相关领域的研究．
The motor unit action potential （MUAP） can provide significant physiological parameters about the neural muscular system. Based on the Blind Source Separation （BSS） methods of Independent Component Analysis （ICA） and Second Order Non-stationary Source Separation （SEONS）, the MUAP detection of surface electromyography （SEMG） at low contraction force is explored. Utilizing the simulated SEMG signals, the performance of ICA algorithm is analyzed and compared with that of the decomposition technique adopting SEONS, the decomposition experiments of real SEMG signals that record at low contraction force （10%MVC） are also done. The experiment results show that ICA and SEONS methods can decompose simulated and recorded SEMG signals effectively, and the performance degrades with increasing noise in the simulated experiments. The appropriate approach is provided to detect the motor unit recruitment, firing and other information through BSS, so the BSS technique can be used for the study about SEMG signals decomposition, MUAP detection and so on.
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2006年第B07期558-563,共6页
Journal of Harbin Engineering University
基金
国家自然科学基金资助项目（60371015）.
关键词
表面肌电信号
盲源分离
独立分量分析
二阶非平稳源分离
运动单位动作电位
surface electromyography
blind source separation
independent component analysis
second order non-stationary source separation
motor unit action potential
