Node Centrality Measures Identify Relevant Structural MRI Features of Subjects with Autism

Area of Science:

• Neuroscience
• Developmental Neuroscience
• Brain Imaging

Background:

• Autism spectrum disorders (ASDs) are complex neurodevelopmental conditions.
• Genetic and environmental factors likely interact to affect early brain development.
• Magnetic resonance imaging (MRI) can detect group-level brain alterations in ASD.

Purpose of the Study:

• To apply a network-based approach to characterize brain structural differences in ASD.
• To identify variations in brain features and their relationships between autistic and typically developing individuals.
• To pinpoint specific features distinguishing these groups.

Main Methods:

• Network-based analysis of structural MRI (sMRI) data from the ABIDE-I database.
• Extraction of 419 features using FreeSurfer software.
• Comparison of network properties (node centrality) between autistic (AUT) and typically developing (TD) groups, using subsampling for class imbalance.

Main Results:

• Significant differences were observed in averaged node centrality measures between the AUT and TD groups.
• A single-node analysis identified specific features that effectively distinguished the two groups.
• The network approach revealed distinct patterns in feature relationships, reflecting underlying structural brain differences.

Conclusions:

• Network analysis provides a powerful method to uncover subtle brain structural differences in ASD.
• Specific brain features and their interconnections are key discriminators between autistic individuals and controls.
• This approach enhances our understanding of the neurobiological underpinnings of autism spectrum disorders.