不同行为状态下大鼠脑电时频变化特性的直方图分析

结合应用多分辨率小波分解方法和直方图参数统计方法 ,分析大鼠脑电信号 (Electroencephalogram,EEG)在不同行为状态下的非稳态时频动态变化特性。利用埋植电极记录自由活动大鼠在清醒期、慢波睡眠期和快动眼睡眠期的皮层 EEG,应用小波变换将 EEG分解成 δ、θ、α和 β四个分量 ,求各分量功率对数值直方图和功率百分比值直方图的均值、方差、偏斜度和峭度。结果表明 :EEG功率对数值的分布比较接近正态分布 ,而多数功率百分比值的分布与正态分布差别显著。单因素方差分析结果显示这些直方图统计参数在不同行为状态之间和不同分解分量之间具有显著差别。 EEG在不同时期的某些特征波 (例如 :慢波睡眠期的 δ波、清醒期和快动眼睡眠期的 θ波等 )使功率对数值分布具有较大的偏斜度值和峭度值。由此可见 ,EEG小波分解分量的直方图参数是一种新的描述EEG动态时频变化特性的定量分析指标

Using the Histogram Analysis Method to Assess the Time-frequency Features of Rat EEG under Different Vigilance States

To investigate the non-stationary time-frequency features in rat Electroencephalogram (EEG) under different vigilance states, the methods of multi-resolution wavelet transform (WT) and statistical histogram analysis were used. EEGs of the freely moving rats were recorded with implanted electrodes under the vigilance states of waking, slow wave sleep (SWS) and rapid eye movement sleep (REM). The EEGs were firstly decomposed into four frequency components of delta, theta, alpha and beta by using multi-resolution wavelet transform. Then, the parameters of mean value, standard deviation, skewness and kurtosis of the logarithm power histograms and the power percentage histograms of each of the frequency components were calculated. The results showed that the distributions of the logarithm power histograms were not quite different from the normal distribution. However, most of the power percentage histograms were significantly different from the normal distribution. The results of one-way ANOVA indicated that there were significant differences in the parameter values of the histograms both among different states and among different frequency components. Moreover, Skewness and kurtosis of the logarithm power histograms of some characteristic waves in EEG, such as delta wave during SWS and theta wave during waking and REM, obtained high values. Thus, the histogram parameters of EEG WT components might become as quantitative measures to describe the dynamic time-frequency features of EEG.