Institute of High Performance Computing (IHPC), Agency for Science, Technology and Research (A*STAR), Singapore.
Singapore Institute for Neurotechnology (SINAPSE), Centre for Life Sciences, National University of Singapore, Singapore.
Hum Brain Mapp. 2018 Sep;39(9):3528-3545. doi: 10.1002/hbm.24192. Epub 2018 Apr 24.
Fronto-parietal subnetworks were revealed to compensate for cognitive decline due to mental fatigue by community structure analysis. Here, we investigate changes in topology of subnetworks of resting-state fMRI networks due to mental fatigue induced by prolonged performance of a cognitively demanding task, and their associations with cognitive decline. As it is well established that brain networks have modular organization, community structure analyses can provide valuable information about mesoscale network organization and serve as a bridge between standard fMRI approaches and brain connectomics that quantify the topology of whole brain networks. We developed inter- and intramodule network metrics to quantify topological characteristics of subnetworks, based on our hypothesis that mental fatigue would impact on functional relationships of subnetworks. Functional networks were constructed with wavelet correlation and a data-driven thresholding scheme based on orthogonal minimum spanning trees, which allowed detection of communities with weak connections. A change from pre- to posttask runs was found for the intermodule density between the frontal and the temporal subnetworks. Seven inter- or intramodule network metrics, mostly at the frontal or the parietal subnetworks, showed significant predictive power of individual cognitive decline, while the network metrics for the whole network were less effective in the predictions. Our results suggest that the control-type fronto-parietal networks have a flexible topological architecture to compensate for declining cognitive ability due to mental fatigue. This community structure analysis provides valuable insight into connectivity dynamics under different cognitive states including mental fatigue.
通过社区结构分析发现，额顶子网能够补偿由于精神疲劳导致的认知能力下降。在这里，我们研究了由于长时间执行认知要求高的任务而导致的静息态 fMRI 网络子网络拓扑结构的变化，以及它们与认知能力下降的关系。由于大脑网络具有模块化组织，因此社区结构分析可以提供有关中尺度网络组织的有价值信息，并作为将标准 fMRI 方法与脑连接组学联系起来的桥梁，脑连接组学用于量化整个大脑网络的拓扑结构。我们开发了模块间和模块内网络指标，以量化子网络的拓扑特征，我们的假设是精神疲劳会影响子网络的功能关系。使用基于正交最小生成树的小波相关和数据驱动的阈值方案构建功能网络，这允许检测具有弱连接的社区。在前任务运行和后任务运行之间，发现了额叶和颞叶子网之间的模块间密度的变化。七个模块间或模块内网络指标，主要在额叶或顶叶子网中，表现出对个体认知能力下降的显著预测能力，而整个网络的网络指标在预测中效果较差。我们的结果表明，控制型额顶网络具有灵活的拓扑结构，可以补偿由于精神疲劳导致的认知能力下降。这种社区结构分析为不同认知状态（包括精神疲劳）下的连接动力学提供了有价值的见解。
