一种基于小波变换的脑电信号癫痫特征波的识别算法

目的通过选取小波变换的多尺度特性算法,能够快速而准确的从脑电图信号中识别癫痫特征波。方法确定癫痫脑电的3种特定的病态波形,选取小波变换的多尺度特性算法,分析处理16导标准脑电信号。结果分离出癫痫特征波,并对特征波进行识别,从而得到对癫痫的诊断;在此基础上将癫痫特征波反映射到16导标准电极,应用相关源电位软件对癫痫灶进行初步定位。结论小波变换的多尺度特性算法可以实现对癫痫脑电信号特征波的自动检测和病灶定位,有助于临床诊断和筛查癫痫。

Analyzing Research of EEG Signal Based on Sparse Representation Model

Objective: To verify that the algorithm can identify EEG signal wave of epilepsy quickly and accurately.Methods: EEG epilepsy identified three specific pathological waveform,to select the wavelet transform algorithm of multi-scale analysis with the standard 16-EEG.Results: isolate the wave of epilepsy and a wave of recognition,resulting in the diagnosis of epilepsy;On this basis,reflect the wave of strikes into epilepsy-16 standard electrode potential source software applications related to the initial positioning of epileptic foci.Conclusion: Multi-scale wavelet transform algorithm can implement the right of epileptic EEG waves of automatic detection and localization,clinical diagnosis and screening contribute to epilepsy.