AI framework improves cross-site MDD identification from fMRI data

Researchers have developed a novel framework for identifying major depressive disorder (MDD) from resting-state functional magnetic resonance imaging (rs-fMRI) data across different sites. This approach addresses challenges posed by inter-site distribution shifts and heterogeneous functional connectivity views by jointly modeling multiple graph representations. The framework utilizes view-specific graph attention networks, dual-stream adaptive fusion, and hyperbolic residual encoding for representation refinement, achieving a mean accuracy of 73.60% and an AUC of 71.90% across seven unlabeled target domains. AI

IMPACT This research could lead to more accurate and generalized AI models for medical diagnosis across diverse datasets.

RANK_REASON The item is an academic paper published on arXiv detailing a new methodology. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• arXiv
• Cauchy--Schwarz alignment
• Granger causality
• graph attention network
• hyperbolic residual encoding
• McDonnell Douglas
• Pearson correlation
• RS-fMRI to Nutrient Shakes

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →