Large-Scale Granger Causal Brain Network based on Resting-State fMRI data

Xuewei Wang,Ru Wang,Fei Li,Qiang Lin,Xiaohu Zhao,Zhenghui Hu
DOI: https://doi.org/10.1016/j.neuroscience.2019.11.006
IF: 3.708
2020-01-01
Neuroscience
Abstract:The causal connections among small-scale regions based on resting-state fMRI data have been extensively studied and a lot of achievements have been demonstrated. However, the causal connection among large-scale regions was seldom discussed. In this paper, we applied global Granger causality analysis to construct the causal connections in the whole-brain network among 103 healthy subjects (33M/66F, ages 20-23) based on a resting-state fMRI dataset. We further explored four large-scale cognitive networks which have been widely known: central executive network (CEN), default mode network (DMN), dorsal attention network (DAN) and salience network (SN). These four cognitive networks are particularly important for understanding higher cognitive functions and dysfunction. Based on the above research, Out-In degree were introduced to identify the driving and driven hubs. Studying the driving and driven hub of brain network is of great significance for assessing the functional mechanism of the brain network. There were 817 directed edges identified as significant among the 8010 possible causal connections; seven driving hubs and ten driven hubs were identified in the whole-brain network. In CEN, dorsolateral prefrontal cortex (DlPFC) and superior parietal cortex (SPC) were the driven and driving hubs, respectively; in DMN, they were posterior cingulate cortex (PCC) and medial prefrontal cortex (MPFC); in DAN, they were frontal eye fields (FEF) and intraparietal sulcus (IPS); and in SN, they were frontoinsular cortex (FIC) and medial frontal cortex (MFC). These findings may provide insights into our understanding of human brain function mechanisms and the diagnosis of brain diseases.
neurosciences
What problem does this paper attempt to address?