小波变换用于MLAEP信号的分析

An Analysis of MLAEP Based on Wavelet Transformation

在现代临床麻醉学上 ,脑电信号分析成为麻醉深度监测的主要手段。其中 ,对于中潜伏期听觉诱发电位 (ML AEP)信号的研究越来越受到重视。本文利用 Daubechies正交紧支撑小波基对不同麻醉深度下的 ML AEP信号进行小波变换 ,然后用顺序前进法对小波系数进行特征选择 ,最后利用神经网络来完成分类。仿真实验表明 ,这种小波变换与特征选择相结合的方法 ,对于 ML AEP信号的分析有着较好的效果。

Electroencephalogram(EEG) has become the main method of monitoring the depth of anesthesia in modern clinic anesthesiology, and the study of middle latency auditory evoked potentials(MLAEP) has been widely supplied. In this paper, wavelet transformation on MLAEP with Daubechies orthonomal bases of compactly supported wavelets has been made, then the sequential floating forward search method was used for feature selection,and finally the artifical neural network was used as a pattern classifier . The simulation results have shown that the combination of wavelet transformation and feature selection was effective in MLAEP analysis.