摘要
针对α稳定噪声下诱发电位(EP)潜伏期延长的检测问题,基于一种韧性的时延估计方法(LMSSETDE),得到EP潜伏期延长的估计值。该方法用滑动平均(MA)滤波器模拟神经通路对EP信号传导的延迟效应。MA滤波器输入信号为潜伏期正常的EP信号,滤波器输出参考信号为待测的潜伏期延长的EP信号,在最小均方Sigmoid误差韧性准则下,自适应得到滤波器的最优权系数,其峰值位置为待测EP信号潜伏期相对于正常EP信号潜伏期的延长值。计算机仿真结果表明:LMSSETDE方法是一种有效的EP潜伏期延长检测方法。
A robust time delay estimation method, referred as least mean square sigmoid error time delay estirnation(LMSSETDD, is used to detect the evoked potential latency delay since the evoked potential signals is with stable noises. A moving average filter is used to model the delay effect of the nerve access conducting the evoked potential signals. The input signal of the filter is the evoked potential signals with normal latency, and the output signal of the filter is the evoked potential signals with abnormal latency to be detected. The optimal filter weight coefficients can be adaptively obtained under the least mean square sigmoid error criteria. The peak location of the weight coefficients is regarded as the relative time delay between the evoked potential signals with normal latency and those with abnormal latency to be detected. Computer simulation studies show that the LMSSETDE method is effective for the detection of evoked potential latency delay.
出处
《上海电机学院学报》
2010年第1期4-8,共5页
Journal of Shanghai Dianji University
基金
中国博士后科学基金项目(20070420602)
上海电机学院科研启动经费项目(10C403)
关键词
信号与信息处理
诱发电位潜伏期
韧性时延估计
Α稳定分布
signal and information processing
evoked potential latency
robust time delayestimation
a stable distribution