异常脑功能网络的构建与分析

发布时间：2018-06-07 08:36

本文选题：阿尔兹海默症 + 脑电信号分析　；参考：《天津大学》2016年硕士论文

【摘要】：脑功能网络研究对于了解大脑信息处理方式以及认知表达的神经生理机制具有重要意义,而现代的脑成像技术和复杂网络理论为脑功能网络的研究提供了必要的基础。本文基于静息态的阿尔兹海默症(AD)脑电信号构建异常脑功能网络,并分析其网络拓扑结构的改变,有助于理解AD的发病机制并为诊断提供辅助。本文应用有限穿越可视图与相空间最近邻点建网方法,构建单导联脑电信号的脑功能网络。首先通过将脑电信号时间序列映射为复杂网络,对复杂网络进行网络拓扑分析,提取其网络的特征参数并以此来表征原始时间序列的动力学特性。研究结果显示,与正常老年对照组相比,AD患者单导联脑功能网络的拓扑出现异常,网络平均最短路径长度与聚类系数发生改变。其中显著性变化主要出现在顶部、枕部和右颞部。本文进一步提取序列间的同步性构建并分析多导联的脑功能网络。通过前面得到的单导联复杂网络聚类系数序列,计算两导信号间的皮尔逊相关系数,以及不同导联间的同步似然(SL)和模糊同步似然(FSL)系数,利用图论理论提取网络拓扑参数,分析四个频带下AD脑功能网络的拓扑结构,发现显著性差异主要发生在alpha频带,AD脑功能网络具有“失连接”现象,即脑区之间的脑功能连接减少,导致功能网络的图指数复杂度降低及小世界属性削弱。此外,对于同步似然(SL)和模糊同步似然(FSL)算法,在分析真实实验数据之前,本文利用Henon耦合模型产生仿真数据,验证了该算法的有效性并比较其优劣。最后,经过多种方法和不同维度的脑功能网络的构建与分析,本文得到了阿尔兹海默症对脑功能网络拓扑结构的影响,发现其网络特征参数的异常改变,可以通过检测网络特征参数的变化对阿尔兹海默症进行诊断辅助与康复评估。
[Abstract]:The study of brain functional networks is of great significance in understanding the neurophysiological mechanisms of brain information processing and cognitive expression. Modern brain imaging technology and complex network theory provide a necessary basis for the study of brain functional networks. In this paper, we construct abnormal brain functional network based on resting Alzheimer's disease (AD) EEG signal, and analyze the change of its network topology, which is helpful to understand the pathogenesis of AD and provide assistance for diagnosis. In this paper, we use the method of constructing the network between the finite traversing visual map and the nearest neighbor point in phase space to construct the functional network of the single lead EEG signal. Firstly, by mapping the EEG time series into complex networks, the network topology of complex networks is analyzed, and the characteristic parameters of the network are extracted to characterize the dynamic characteristics of the original time series. The results showed that the topology of single-lead brain functional network in AD patients was abnormal, the average shortest path length and clustering coefficient of AD patients were changed. Significant changes occurred at the top, occipital and right temporal regions. In this paper, we further extract the synchronism between sequences to construct and analyze the multi-lead brain functional network. The Pearson correlation coefficient between the two conductance signals, the synchronous likelihood SLS and the fuzzy synchronous likelihood FSLs are calculated by the clustering coefficient sequence of the single lead complex network, and the topological parameters of the network are extracted by the graph theory. After analyzing the topological structure of AD brain functional network in four bands, it was found that the significant difference occurred mainly in the "disconnection" of AD brain functional network in alpha band, that is, the reduction of brain functional connection between brain regions. It leads to the reduction of graph exponential complexity and the weakening of small world properties. In addition, for the synchronous likelihood (SL) and fuzzy synchronous likelihood (FSL) algorithms, before analyzing the real experimental data, the Henon coupling model is used to generate the simulation data, which verifies the effectiveness of the algorithm and compares its merits and demerits. Finally, through the construction and analysis of various methods and different dimensions of brain functional network, the influence of Alzheimer's disease on the topological structure of brain functional network is obtained, and the abnormal changes of network characteristic parameters are found. Diagnostic aids and rehabilitation assessments can be performed for Alzheimer's disease by detecting changes in network characteristic parameters.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：R749.16;O157.5

【相似文献】