基于静息态脑电的心因性非癫痫性发作患者脑功能网络分析及分类识别研究  

Brain Function Network Analysis and Recognition for Psychogenic Non-epileptic Seizures Based on Resting State Electroencephalogram

作　　者：王振宇[1,2] 薛青[3] 熊秀春[1] 李沛洋[1] 田春阳[1] 符策红[2] 王玉平[3] 尧德中[1] 徐鹏[1] 

机构地区：[1]电子科技大学神经信息教育部重点实验室,成都610054 [2]海南软件职业技术学院信息管理系,琼海571400 [3]首都医科大学宣武医院神经内科,北京100053

出　　处：《生物医学工程学杂志》2015年第1期8-12,共5页Journal of Biomedical Engineering

基　　金：国家自然科学基金资助项目(61175117,31070881,31100745);新世纪优秀人才项目资助(NCET-12-0089);863重点项目资助(2012AA011601);北京市科技计划项目资助(Z101107050210015)

摘　　要：研究表明,神经精神疾病患者的临床表现与其大脑功能网络连接异常是相关的。心因性非癫痫(PNES)是一种心理疾病,不具有特征性的癫痫放电表现,是临床诊断上的一个难点。本文基于采集的脑电图(EEG)信号,利用网络分析方法,发现PNES患者前额与枕顶脑区之间的网络连接强度较正常组减弱。并且,将网络属性作为特征,利用线性判别分析(LDA)可对PNES患者和对照组获得85%的分类准确率,为临床诊断提供具有实际意义的价值信息。Studies have shown that the clinical manifestation of patients with neuropsychiatric disorders might be related to the abnormal connectivity of brain functions.Psychogenic non-epileptic seizures（PNES）are different from the conventional epileptic seizures due to the lack of the expected electroencephalographically epileptic changes in central nervous system,but are related to the presence of significant psychological factors.Diagnosis of PNES remains challenging.We found in the present work that the connectivity between the frontal and parieto-occipital in PNES was weaker than that of the controls by using network analysis based on electroencephalogram（EEG）signals.In addition,PNES were recognized by using the network properties as linear discriminant nalysis（LDA）input and classification accuracy was 85%.This study may provide a feasible tool for clinical diagnosis of PNES.

关 键 词：脑电图 心因性非癫痫 网络分析 线性判别分析 

分 类 号：R741[医药卫生—神经病学与精神病学]