颞叶癫痫全脑结构和功能连接模式的研究

项目名称： 颞叶癫痫全脑结构和功能连接模式的研究

项目编号： No.81471251

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 医药、卫生

项目作者： 邱士军

作者单位： 广州中医药大学

项目金额： 80万元

中文摘要： 颞叶癫痫是一种最常见的难治性、局灶性癫痫，并常与多种认知功能障碍有关。虽然海马硬化是其主要的病理特征，但是越来越多的结构和功能MRI研究提示颞叶癫痫是一种全脑性的网络疾病。但是传统的研究方法很难获得颞叶癫痫全脑的结构和功能连接模式，其次以往研究所采用的组级别的统计方法不能在个体水平评估结构和功能连接的鉴别能力。当前机器学习方法已经广泛的应用于磁共振数据分析，因为它能够从影像数据中提取出新的信息和感兴趣的模式，找到基于全脑影像数据的生物学标记，并从个体水平区别病人与正常人。本项目拟在我们前期对颞叶癫痫患者和正常人脑白质部分各向异性值进行分类的研究基础上，利用机器学习方法来研究颞叶癫痫全脑结构与功能连接模式，通过分析结构和功能连接模式的相关性以及两者与临床各指标的相关性，以期能够加深对于颞叶癫痫的病理生理机制的理解和认识，为颞叶癫痫的临床诊断、药物治疗、术前及术后评估提供较为系统的科学依据。

中文关键词： 颞叶癫痫；结构连接；功能连接；机器学习

英文摘要： Temporal lobe epilepsy, often associated with cognitive impairments, is the most common type of refractory focal epilepsy. Although hippocampal sclerosis is the hallmark of this condition, a growing body of structural and functional MRI studies demonstrates that temporal lobe epilepsy is potentially a brain network dysfunction. However,traditional research methods are difficult to examine stuctural and functional patterns on a whole-brain scale. In addition, traditional group-level statistical methods do not provide a mechanism for evaluating the discriminative power of the identified structural and functional connections at the individual level. In recent years, machine learning approaches have been increasing used for brain image analysis because they are capable of extracting additional information and stable patterns from neuroimaging data, finding significant whole-brain neuroimaging-based biomarkers and identifying patients from controls at individual subject levels. This research project is based on our previous findings that fractional anisotropy (FA) values of white matter can reliably differentiate temporal lobe epilepsy patients from healthy volunteers, applying machine learning approaches to whole-brain structural and functional connectivity analysis in temporal lobe epilepsy, investigating the correlation between this two patterns and exploring the correlation between the two pattern and the clinical index, aiming to improve the understanding of pathophysiological mechanisms of the temporal lobe epilepsy and provide a more scientific basis for diagnosis, treatment, preoperative and postoperative evaluation of the temporal lobe epilepsy.

英文关键词： temporal lobe epilepsy;structural connectivity;functional connectivity;machine learning