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

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Network Science

Background:

• Resting-state fMRI studies have focused on small-scale brain regions, leaving large-scale causal connections under-explored.
• Understanding large-scale brain networks is crucial for higher cognitive functions and diagnosing brain dysfunction.

Purpose of the Study:

• To apply global Granger causality analysis to construct whole-brain causal networks in healthy adults.
• To identify driving and driven hubs within four major cognitive networks: CEN, DMN, DAN, and SN.
• To investigate the functional mechanisms of brain networks and their relevance to brain disease diagnosis.

Main Methods:

• Global Granger causality analysis applied to resting-state fMRI data from 103 healthy subjects (ages 20-23).
• Construction of directed causal connections within the whole-brain network.
• Identification of driving and driven hubs using Out-In degree metrics within CEN, DMN, DAN, and SN.

Main Results:

• 817 significant directed causal edges identified out of 8010 possible connections.
• Seven driving hubs and ten driven hubs identified in the whole-brain network.
• Specific hubs identified in each network: CEN (SPC-driving, DlPFC-driven), DMN (MPFC-driving, PCC-driven), DAN (IPS-driving, FEF-driven), SN (MFC-driving, FIC-driven).

Conclusions:

• The study successfully mapped large-scale causal brain networks and identified critical driving and driven hubs.
• Findings provide insights into the functional mechanisms of human brain networks.
• The identified network patterns may aid in the diagnosis and understanding of brain diseases.