aiqbench analysis
HyperAMS-Net presents a novel deep learning architecture leveraging hypergraph attention and adaptive multi-scale convolution to tackle inter-subject heterogeneity in neuroimaging data. By modeling higher-order dependencies among feature representations, the framework achieves state-of-the-art accuracy across benchmarks for autism, depression, and Alzheimer's disease. This work advances machine learning applications in medical imaging by demonstrating how complex graph-based structures can effectively capture intricate patterns in functional connectivity and morphological data.
HyperAMS-Net: Adaptive Multi-Scale Spatial Hypergraph Network for Brain Disorder Classification
arXiv:2609.19755v1 Announce Type: cross Abstract: Accurate classification of brain disorders from neuroimaging data remains challenging because of substantial inter-subject heterogeneity and the complex multi-scale patterns present in functional connectivity and morphological representations. To address these challenges, we propose HyperAMS-Net, a deep learning framework for brain disorder classification using neuroimaging representations derived from resting-state functional MRI or structural MRI. HyperAMS-Net integrates adaptive multi-scale convolution, hypergraph attention, spatial-channel attention, and adaptive feature fusion. Specifically, adaptive multi-scale convolution learns data-driven weights over multiple receptive fields to capture complementary patterns at different scales. Hypergraph attention models higher-order dependencies among learned feature representations through node--hyperedge--node message passing, while spatial-channel attention enhances discriminative feature learning. Adaptive feature fusion further aggregates complementary information across parallel network branches. HyperAMS-Net is evaluated on three benchmark datasets spanning distinct brain disorders: ABIDE for autism spectrum disorder, REST-meta-MDD for major depressive disorder, and ADNI for Alzheimer's disease, using 5-fold stratified cross-validation. HyperAMS-Net achieves state-of-the-art performance across all evaluated datasets, attaining the highest accuracy and AUC among the compared methods. Ablation studies further demonstrate the contribution of each proposed component, with the largest performance degradation observed when hypergraph attention is removed.
Abstract:Accurate classification of brain disorders from neuroimaging data remains challenging because of substantial inter-subject heterogeneity and the complex multi-scale patterns present in functional connectivity and morphological representations. To address these challenges, we propose HyperAMS-Net, a deep learning framework for brain disorder classification using neuroimaging representations derived from resting-state functional MRI or structural MRI. HyperAMS-Net integrates adaptive multi-scale convolution, hypergraph attention, spatial-channel attention, and adaptive feature fusion. Specifically, adaptive multi-scale convolution learns data-driven weights over multiple receptive fields to capture complementary patterns at different scales. Hypergraph attention models higher-order dependencies among learned feature representations through node--hyperedge--node message passing, while spatial-channel attention enhances discriminative feature learning. Adaptive feature fusion further aggregates complementary information across parallel network branches. HyperAMS-Net is evaluated on three benchmark datasets spanning distinct brain disorders: ABIDE for autism spectrum disorder, REST-meta-MDD for major depressive disorder, and ADNI for Alzheimer's disease, using 5-fold stratified cross-validation. HyperAMS-Net achieves state-of-the-art performance across all evaluated datasets, attaining the highest accuracy and AUC among the compared methods. Ablation studies further demonstrate the contribution of each proposed component, with the largest performance degradation observed when hypergraph attention is removed.
Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
arXiv:2609.19755 [eess.IV]
(or arXiv:2609.19755v1 [eess.IV] for this version)
https://doi.org/10.48550/arXiv.2609.19755 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 289 words
Related coverage
Learning Multi-Scale Hypergraph for High-Order Brain Connectivity Analysis
arXiv · Jun 3, 2026
Does preservation make sense before we know how to revive?
LessWrong · Jun 15, 2026
Google DeepMind Ships Three Physical AI Models For Whole Body Control, Dexterity And Multi Robot Collaboration
MarkTechPost · Jul 30, 2026
Step-adaptive multimodal fusion network with multi-scale cloud feature learning for ultra-short-term solar irradiance forecasting
arXiv · Jun 6, 2026