论文部分内容阅读
目的 :评价滑行窗口技术分析脑电诱发电位的能力。方法 :将具有一定宽度的时间窗口延时间轴滑行 ,计算该窗口内的脑电电位平均值 ,再与对照窗口进行统计比较 ,以检验诱发电位是否具有统计显著性。利用该方法分析随机产生的模拟数据 ,计算在指定单次检验阈值下 ,多次统计比较导致显著性差异点连续出现的几率 ,以确定可使整体α值小于 0 .0 5的cluster大小。为检验该方法的有效性 ,在 14名健康右利手志愿者右手中指给予痛或非痛电刺激 ,记录EEG信号并采用上述技术加以分析。结果 :在整体α值确定的前提下 ,作为显著性判据的cluster大小随单次检验阈值与窗宽的增加而增大。依据上述方法分析真实EEG数据 ,确定了体感与痛觉诱发电位波形中具有统计学意义的成分 ,以及两种波形之间的显著性差异。结论 :滑行窗口技术可有效地用于分析脑电诱发电位。
OBJECTIVE: To evaluate the ability of gliding window technique to analyze EEG. Methods: The time window with a certain width was glided on the time-lag axis, the average of the EEG in this window was calculated, and then compared with the control window to test whether the evoked potential was statistically significant. This method is used to analyze the randomly generated simulation data to calculate the probability of continuous occurrence of significant differences under multiple statistical comparisons under the specified single test threshold to determine the cluster size that can make the overall α smaller than 0.05. To test the effectiveness of the method, pain or non-painful electrical stimulation was given to the right middle finger of 14 healthy right hand volunteers, EEG signals were recorded and analyzed using the above techniques. Results: Under the premise of determining the overall α value, the cluster size as a significant criterion increases with the increase of single test threshold and window width. Based on the above analysis of real EEG data, statistically significant components of somatosensory and pain evoked potential waveforms and significant differences between the two waveforms were identified. Conclusion: Glide window technique can be effectively used to analyze EEG.