Studying the human brain anatomical network via diffusion-weighted MRI and Graph Theory

Area of Science:

• Neuroscience
• Network Science
• Brain Imaging

Background:

• Understanding the human brain's anatomical network is crucial for neuroscience.
• Diffusion-weighted Magnetic Resonance Imaging (DW-MRI) provides data for network estimation.
• Complex network analysis offers insights into brain structure and function.

Purpose of the Study:

• To estimate anatomical connection probabilities (ACP) between 90 brain regions.
• To model the brain as a weighted graph and analyze its complex network properties.
• To identify critical brain areas based on network vulnerability and centrality.

Main Methods:

• Estimated ACP from DW-MRI data for 20 healthy subjects.
• Modeled the brain as a non-directed weighted graph using the ACP matrix.
• Analyzed network properties: small-world attributes, efficiency, centrality, and motifs.

Main Results:

• All brain networks exhibited small-world and broad-scale characteristics.
• Brain networks showed higher local efficiency and lower global efficiency than random networks.
• Identified critical areas: putamens, precuneus, insulas, superior parietals, and superior frontals.
• Some areas demonstrated negative vulnerability, suggesting evolutionary importance.

Conclusions:

• Human brain anatomical networks possess small-world properties and specific efficiency profiles.
• Key brain regions were identified through network analysis, highlighting their structural importance.
• Motif analysis revealed hierarchical structural organization within the brain's anatomical network.