基于偏相关系数的脑功能连接分析

摘要

The human brain is an information processing system of high complexity with complex structure and various functions which is mainly dependent on the network composed of the interaction between brain regions.This paper adopts partial correlation coefficient analysis with a threshold setting to get a further introduction called node degree.Its value can tell us the connection strength between nodes,we used it to analyze 273 brain magnetic signal channels (MEG)under resting state in 24 healthy subjects and 13 cases of mental patients.Explored the difference between their brain function bindings.Studies have shown that patients with mental disease have more connection between brain areas,there is a significant difference in the frontal and temporal lobes and the limbic portion of normal subjects.We suppose that these abnormalities are related to schizophrenia.%大脑是一个具有高度复杂性的信息处理系统,拥有丰富且复杂的结构与功能,其功能的执行主要依赖于脑功能区之间的相互作用所构成的网络实现.研究采用偏相关系数分析方法,在阈值的设定研究基础上,进一步引入节点度的概念,用于反馈节点之间连接强度的大小.对13例精神分裂症患者和24例正常人在静息态下273个通道的脑磁(MEG)信号进行分析,探索正常组与患者组之间脑功能连接的差异性.研究表明精神病患者在大部分脑区有显著的功能连接,在额叶和颞叶以及脑边缘部分与正常人有显著性差异,推测这些异常连接区域有可能与精神分裂症的病理有关.