多分辨率小波信号分解用于大鼠睡眠纺锤波的分析

USING THE MULTI-RESOLUTION WAVELET DECOMPOSITION TO ANALYZE THE SLEEP SPINDLE IN RAT EEG

本研究首先设计了慢波睡眠期脑电信号的合成仿真信号 ,对小波基函数进行了选择 ,结果证明Coiflet 5阶小波变换对大鼠慢波睡眠期EEG信号具有较好的分解结果。据此 ,应用多分辨率小波分析法设计了提取睡眠纺锤波的算法 ,并利用该算法对安定用药后和睡眠剥夺后大鼠慢波睡眠期纺锤波的持续时间和能量变化进行了分析 ,结果表明 :安定具有延长慢波睡眠期纺锤波持续时间的作用 ,而睡眠剥夺可以增加慢波睡眠期纺锤波的能量。这些结果说明 ,小波分析算法可以提供功率谱分析无法表现的时频信息。

An algorithm was designed to automatically analyze the sleep spindles in rat sleep EEG based on the multi-resolution wavelet decomposition. This algorithm was used to evaluate the effects of the diazepam and the sleep deprivation on the sleep spindles of rat EEG. Firstly, a simulating EEG signal with sleep spindles was synthesized and used to select a more suitable wavelet function for decomposing sleep spindles from EEG. The results showed that the best result was obtained using Coiflet wavelet. Then, this wavelet function was used to establish the algorithm to extract the sleep spindles from EEG with a threshold method. The EEGs were recorded from freely moving rats with implanted electrodes following the applications of diazepam or following sleep deprivations. The results showed that the duration of sleep spindle was significantly increased in the diazepam sleep but not in the recovery sleep following sleep deprivation, however, the power of sleep spindle in the recovery sleep was significantly increased. Therefore, the method of wavelet decomposition can provide more information in detail on both time and frequency fields than the traditional power spectrum.