基于经验模态分解的颈部表面肌电信号特征提取

Characteristic Extraction of Surface Electromyography Signals based on Empirical Mode Decomposition

通过SEMG的线性与非线性指标分析评估被试者在伏案工作过程中的颈部肌肉疲劳程度.采用经验模态分解EMD将原始信号中的强背景噪声分解至各个固有模态函数IMF,达到降噪的作用.在此基础上,分别对SEMG的线性、非线性指标进行了分析.结果表明:随着低头伏案学习时间的增加,被试者的肌肉疲劳程度不断加重,线性和非线性指标随时间都有较明显的线性变化规律.此项研究结果表明长期伏案工作者中的颈椎病患者通常易发生低节段颈间盘退变,同医学影像学中核磁检测结果一致.因此应用SEMG信号的时频分析评估颈部肌肉疲劳状态是可靠且可行的.

Fatigue degree of cervical muscles during subjects＇ working with bowing heads is evaluated by linear and none-linear characteristic extraction of surface electromyography（SEMG）.The original signal containing noise and artificial contamination is decomposed into several intrinsic mode functions（IMF） by the method of empirical mode decomposition（EMD）,so the noise is reduced.Then the linear and none-linear characteristics of SEMG are analyzed.The experimental results show these characteristics have obvious linear variation with the time.When the working time with bowing head is increased,the fatigue degree of muscle is increased.So the cervical spondylosis usually occurs at lower cervical segment,which is accordant with the measuring result of magnetic resonance imaging.Therefore,the evaluation for fatigue degree of cervical muscles by characteristic extraction of SEMG is feasible and reliable.