Personalized Structure Preservation Based Graph Neural Network via Connection Interaction and Refinement for Autism Spectrum Disorder Diagnosis.

Researchers developed a new computer program called PSP-GNN that can help diagnose autism by looking at brain scans. The program is special because it creates a unique brain map for each person, rather than using the same approach for everyone. It also looks at how different brain areas work together, both directly and indirectly. The researchers say their method works well and finds important brain areas that doctors already know are related to autism.

This study presents PSP-GNN, a personalized Graph Neural Network approach for autism spectrum disorder (ASD) diagnosis using functional brain networks. The method addresses limitations in existing approaches by implementing three key innovations: personalized structure preservation that accounts for individual brain connectivity differences, a connection interaction module that analyzes both direct and indirect brain region relationships, and flexible brain region identification without rigid thresholds. The researchers report that PSP-GNN demonstrates effectiveness in ASD diagnosis and identifies brain regions consistent with established medical knowledge, suggesting potential utility as clinical biomarkers. However, specific sample sizes, comparison groups, and validation metrics are not provided in the available abstract.

This computational approach may contribute to future autism diagnostic tools by providing personalized brain network analysis. However, clinical validation, regulatory approval, and integration with existing diagnostic practices would be required before clinical implementation. The identification of potential biomarkers could support objective diagnostic processes.

Sample size, participant characteristics, validation methods, and comparison with existing diagnostic tools are not reported. The study appears to be primarily a technical development with unclear clinical validation. Specific accuracy metrics and real-world applicability remain unclear from the available information.

Graph Neural Networks (GNNs) have garnered widespread recognition in the identification of Autism Spectrum Disorder (ASD) owing to their remarkable adaptability to irregular patterns of Functional Brain Networks (FBNs). However, current methods for constructing FBNs generally employ a uniform modeling strategy to process neuroimaging data from different subjects, which fail to consider the heterogeneity of functional connectivity patterns among individuals adequately. In addition, existing methods tend to excessively focus on directly connected brain Regions of Interest (ROIs) when analyzing brain networks, underestimat\ing the importance of indirectly connected brain ROIs. At the same time, conventional approaches for identifying crucial brain regions may miss vital regions due to rigid threshold constraints.

To address these issues, we propose Personalized Structure Preservation based GNN (PSP-GNN) for ASD diagnosis, which incorporates three aspects: 1) A personalized structure preservation strategy that constructs individualized brain networks by accounting for subject-specific variations; 2) A connection interaction-aware module designed to characterize interactions between directly and indirectly connected brain regions, providing comprehensive brain network representations; 3) A flexible brain region refinement technique based on Bernoulli sampling, which identifies salient brain regions without relying on pre-defined thresholds. Experimental results demonstrate the effectiveness of PSP-GNN in ASD diagnosis, highlighting its potential as a robust tool for future ASD diagnosis applications that combine FBNs and GNNs. Notably, the critical brain regions identified by PSP-GNN are consistent with established medical knowledge, suggesting their utility as potential biomarkers for clinical ASD diagnosis.