Geib Benjamin R, Stanley Matthew L, Dennis Nancy A, Woldorff Marty G, Cabeza Roberto
Department of Psychology and Neuroscience, Duke University, Durham, North Carolina.
Department of Psychology, Pennsylvania State University, University Park, Pennsylvania.
Hum Brain Mapp. 2017 Apr;38(4):2242-2259. doi: 10.1002/hbm.23518. Epub 2017 Jan 23.
Multivariate functional connectivity analyses of neuroimaging data have revealed the importance of complex, distributed interactions between disparate yet interdependent brain regions. Recent work has shown that topological properties of functional brain networks are associated with individual and group differences in cognitive performance, including in episodic memory. After constructing functional whole-brain networks derived from an event-related fMRI study of memory retrieval, we examined differences in functional brain network architecture between forgotten and remembered words. This study yielded three main findings. First, graph theory analyses showed that successfully remembering compared to forgetting was associated with significant changes in the connectivity profile of the left hippocampus and a corresponding increase in efficient communication with the rest of the brain. Second, bivariate functional connectivity analyses indicated stronger interactions between the left hippocampus and a retrieval assembly for remembered versus forgotten items. This assembly included the left precuneus, left caudate, bilateral supramarginal gyrus, and the bilateral dorsolateral superior frontal gyrus. Integrative properties of the retrieval assembly were greater for remembered than forgotten items. Third, whole-brain modularity analyses revealed that successful memory retrieval was marginally significantly associated with a less segregated modular architecture in the network. The magnitude of the decreases in modularity between remembered and forgotten conditions was related to memory performance. These findings indicate that increases in integrative properties at the nodal, retrieval assembly, and whole-brain topological levels facilitate memory retrieval, while also underscoring the potential of multivariate brain connectivity approaches for providing valuable new insights into the neural bases of memory processes. Hum Brain Mapp 38:2242-2259, 2017. © 2017 Wiley Periodicals, Inc.
神经影像数据的多变量功能连接分析揭示了不同但相互依存的脑区之间复杂、分布式相互作用的重要性。近期研究表明,功能性脑网络的拓扑特性与认知表现的个体和群体差异相关,包括情景记忆方面。在构建了基于事件相关功能磁共振成像记忆检索研究的全脑功能网络后,我们研究了遗忘单词和记忆单词之间功能脑网络结构的差异。本研究得出了三个主要发现。第一,图论分析表明,与遗忘相比,成功记忆与左海马体连接图谱的显著变化以及与大脑其他区域有效通信的相应增加有关。第二,双变量功能连接分析表明,对于记忆项目与遗忘项目,左海马体与一个检索组件之间的相互作用更强。这个组件包括左楔前叶、左尾状核、双侧缘上回和双侧背外侧额上回。记忆项目的检索组件的整合特性比遗忘项目更大。第三,全脑模块化分析表明,成功的记忆检索与网络中模块化结构的分离程度降低存在微弱的显著关联。记忆条件和遗忘条件之间模块化降低的幅度与记忆表现相关。这些发现表明,在节点、检索组件和全脑拓扑层面整合特性的增加有助于记忆检索,同时也强调了多变量脑连接方法在为记忆过程的神经基础提供有价值新见解方面的潜力。《人类大脑图谱》38:2242 - 2259, 2017。© 2017威利期刊公司。