Learning Multi-Scale Hypergraph for High-Order Brain Connectivity Analysis
Researchers have developed a new framework called MuHL to analyze brain connectivity by modeling higher-order interactions between brain regions. Unlike previous methods that focus on pairwise connections or predefined hyperedges, MuHL dynamically learns these complex relationships across multiple resolutions. Experiments show MuHL improves the classification of neurodegenerative diseases like Alzheimer's and Parkinson's, and helps identify key brain regions associated with disease progression. AI
IMPACT This framework could lead to earlier and more accurate diagnoses of neurodegenerative diseases by improving the analysis of complex brain network interactions.