文章摘要

文章信息/Info

Title:: Classification of Epileptic EEG Signals based on Principal Component Analysis and Localization of Epileptic Foci

作者:: 李冬梅1; 尔西丁·买买提1; 杨日东3; 陈子怡4; 田翔华1; 董楠3; 张洋2; 周毅1; 3△; 1. 新疆医科大学研究生学院，新疆乌鲁木齐 830011；2.新疆医科大学第一附属医院神经内科，新疆乌鲁木齐 830011；3.中山大学中山医学院生物医学工程系，广东广州 510080；4.中山大学附属第一医院神经内科，广东广州 510080

Author(s):: LI Dongmei1; ALCITIN Mamat1; YANG Ridong3; CHEN Ziyi4; TIAN Xianghua1; DONG Nan3; ZHANG Yang2; ZHOU Yi1; 3; 1.Postgraduate college，Xinjiang Medical University，Urumqi 830011，Xinjiang，China；2.Neurology,The First Affiliated Hospital of Xinjiang University,Urumqi 830011, Xinjiang；3.Department of Biomedical Engineering,Medical College,Sun Yat-sen University,Guangzhou 510080，Guangdong，China;4.Neurology,The First Affiliated Hospital of Sun Yat-sen University,Guangzhou 510080，Guangdong

Keywords:: Epilepsy; Epileptic focus; Principal component analysis; Random forest; Medical reference range

摘要:: 为了更好的对致痫灶进行准确定位，提出了一种基于PCA（主成分分析）的定位方法。针对非线性动力学方法从不同角度提取癫痫脑电信号特征,首先采用主成分分析对高维特征向量进行降维处理,用随机森林进行分类；随后利用医学参考值范围找出各导联的差异变化，进而实现对致痫灶的初步定位。

Abstract:: In order to better locate seizure focus, a PCA (principal component analysis，PCA) -based localization method was proposed. In order to extract the characteristics of EEG from different angles, PCA was used to reduce the dimensionality of high-dimensional eigenvectors, and then to classify them with random forest. Then, the medical reference range was used to find out the difference of lead change and the preliminary localization of seizure was focused on.