Single-subject morphological brain networks: connectivity mapping, topological characterization and test-retest...

Area of Science:

• Neuroimaging
• Network Neuroscience
• Brain Anatomy

Background:

• Structural MRI traditionally analyzes local brain features.
• Coordination patterns of these features across regions remain poorly understood.
• Morphological brain networks offer a novel approach to study brain organization.

Purpose of the Study:

• To construct and analyze individual-level morphological brain networks.
• To evaluate the topological organization and test-retest reliability of these networks.
• To investigate the impact of analytical choices (spatial smoothing, parcellation, network type) on network properties and reliability.

Main Methods:

• 57 healthy participants underwent two MRI scan sessions.
• Individual morphological brain networks were built using gray matter volume similarity (Kullback-Leibler divergence).
• Graph theory metrics assessed network topology and reliability (intra-class correlation).

Main Results:

• Morphological brain networks showed high reproducibility, especially for interhemispheric connections.
• Networks exhibited nonrandom small-world topology, high efficiency, and modularity.
• Spatial smoothing significantly influenced network characterization but improved reliability; nodal centralities correlated positively with reliability.

Conclusions:

• Individual morphological network analysis is a valid and reliable method for characterizing human brain structure.
• This approach facilitates understanding intersubject variability in behavior and function.
• Morphological network biomarkers hold promise for diagnosing brain disorders.