癫痫EEG相空间变换、混沌特征指标和综合分析模型的研究

项目名称： 癫痫EEG相空间变换、混沌特征指标和综合分析模型的研究

项目编号： No.61263011

项目类型： 地区科学基金项目

立项/批准年度： 2013

项目学科： 自动化技术、计算机技术

项目作者： 周毅

作者单位： 新疆医科大学

项目金额： 45万元

中文摘要： 癫痫发病过程中脑电信号（EEG）变化是其脑内复杂活动的动力学反映，申请人在前期研究中观察到癫痫病人EEG相空间的变化，成功构建其发作时吸引子，对多个混沌特征指标进行了预测研究并获得良好结果。本项目拟在前期工作基础上，结合更充分的癫痫头皮和皮层EEG数据，对癫痫发病过程的相空间变换及大脑混沌系统的状态转变的分叉过程进行研究；建立更准确合理的浮动变化相空间，以支持李雅普诺夫指数、相关维数、近似熵及复杂度等相关混沌特征指标的计算分析，并引入特征指标的变化率等新特征指标；在混沌特征指标对不同类型癫痫EEG有效性研究的基础上，建立个性化的加权评价指标体系和综合分析模型，进而支持癫痫病人病灶区定位和发作预警机制的研究，并进行计算机模拟和临床实验验证。本项目在获得癫痫发病机理更深入的动力学解释及脑神经系统混沌动力学新成果的同时，将为后续开发癫痫病灶定位、发作预警及治疗系统提供理论和应用方法。

中文关键词： 癫痫；脑电信号；非线性动力学；预测；定位

英文摘要： The dynamic change of electroencephalogram (EEG) in process of epileptic seizure is the reflection of dynamic activities in patients' brain. The applicant reconstructed the attractor of the epileptic patients' EEG signal, observed the transformation of phase space, studied 3 chaos's characteristic indexes in pilot studies and got good result. In this project, we will use adequate EEG to study the phase space transformation and the fork process in the brain of chaotic system state transition in the progress of epileptic seizure. We will construct a more reasonable floating phase space to support the chaotic characteristic indexes including Lyapunov exponent, Correlation dimension, Approximate entropy, related complexity and introduce the change rate of those dynamics which is a new feature index. We study the chaos's characteristic index from the different types of epilepsy, construct individual weighted index assess system and a comprehensive analysis model. This research will do computer simulation and clinical trials to support the research on epileptic focus localization and seizure prediction. This project will get a more reasonable dynamics explain of epilepsy pathogenesis mechanism and newly theoretical results of dynamics system of brain. Meanwhile, it will make a solid foundation of theory and method for

英文关键词： epilepsy；EEG；Nonlinear dynamics；prediction；location