基于希尔伯特边际谱和极限学习机的癫痫脑电信号分类  

Epileptic EEG signal classification based on Hilbert marginal spectrum and extreme learning machine

作　　者：火元莲[1] 陈萌萌 郑海亮 连培君 张健 HUO Yuanlian;CHEN Mengmeng;ZHENG Hailiang;LI AN Peijun;ZHANG Jian(College of Physics and Electronic Engineering,Northwest Normal University,Lanzhou,Gansu 730070,China)

机构地区：[1]西北师范大学物理与电子工程学院,甘肃兰州730070

出　　处：《光电子．激光》2021年第10期1083-1091,共9页Journal of Optoelectronics·Laser

基　　金：国家自然科学基金(61561044)资助项目。

摘　　要：提出了一种基于希尔伯特边际谱和极限学习机相结合的癫痫脑电信号分类方法。首先将脑电信号进行经验模态分解,对前5个本征模态函数进行希尔伯特变换,得到其希尔伯特边际谱;然后将希尔伯特边际谱的Shannon熵、Renyi熵和Tsallis熵,以及5个不同频段节律信号的能量作为有效特征输入极限学习机进行分类。实验结果表明,本文方法对癫痫信号的分类准确率达到了99.8%,相比其它分类方法具有更高的检测精度和运算速度,对癫痫发作的实时检测具有潜在的应用价值。This paper presents a classification method of epilepsy electroencephalogram(EEG)signals based on the combination of Hilbert marginal spectrum and extreme learning machine.Firstly,the empirical mode decomposition of EEG signals is carried out,and the Hilbert marginal spectrum is obtained by applying the Hilbert transform to the first 5 intrinsic mode function;Then Shannon entropy,Renyi entropy and Tsallis entropy of Hilbert marginal spectrum as well as the energy of 5 different frequency sub-band rhythm signals were input into the extreme learning machine as effective characteristics for classification.The experimental results show that the classification accuracy of the epileptic signal in this paper reaches 99.8%,which is higher than other classification methods in detection accuracy and computing speed,and has potential application value in real-time detection of epileptic seizures.

关 键 词：癫痫脑电(electroencephalogram EEG)信号 希尔伯特边际谱 极限学习机 谱熵 子带能量 

分 类 号：TN911.7[电子电信—通信与信息系统]