基于脑电信号的癫痫疾病智能诊断与研究  

Intelligent Diagnosis and Research of Epileptic Diseases Based on EEG Signals

作　　者：柳长源[1] 张付浩 韦琦[1] LIU Chang-yuan;ZHANG Fu-hao;WEI Qi(School of Electrical and Electronic Engineering,Harbin University of Science and Technology,Harbin 150080,China)

机构地区：[1]哈尔滨理工大学电气与电子工程学院,黑龙江哈尔滨150080

出　　处：《哈尔滨理工大学学报》2018年第3期91-98,共8页Journal of Harbin University of Science and Technology

基　　金：黑龙江省自然科学基金(F2016022)

摘　　要：针对医疗诊断中癫痫脑电信号分类准确率低、分类类别少的问题,依据粒子群算法和支持向量机理论,提出了一种基于粒子群算法优化支持向量机参数的信号分类检测技术。首先利用小波分析对脑电信号进行5层分解与重构,然后提取含有癫痫特征频率的3、4、5层重构信号的波动系数和近似熵等特征,计算不同状态不同尺度的脑电信号能量,根据不同状态不同尺度的能量分布,调整特征向量的系数。最后使用粒子群算法优化的支持向量机对脑电信号进行分类。实验结果表明,本文提出的方法可以正确识别健康、癫痫发作间期、癫痫发作期3种类型脑电信号,最终的识别率可以达到99.83%。Aiming at the problem of low accuracy and classification of epileptic EEG in medical diagnosis,a signal classification and detection technique based on particle swarm optimization（PSO） was proposed to optimize the support vector machine（SVM） based on the theory of particle swarm optimization and support vector machine（SVM）. Firstly, the EEG signals were decomposed and reconstructed by wavelet analysis. Secondly, the coefficients of fluctuation and approximate entropy of the reconstructed signals containing the functional parameters of epilepsy were extracted. Finally,The support vector machine（SVM） optimized by particle swarm optimization（PSO） is used to classify the EEG signals. The experimental results show that the this method can correctly identify three types of EEG signals in healthy,interictal epilepsy and epileptic seizures,the final recognition rate can reach99. 83%.

关 键 词：癫痫脑电信号 波动系数 近似熵 粒子群算法 支持向量机 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术] R318.04[自动化与计算机技术—计算机科学与技术]