Research map / Vision-language and multimodal / Cross-modal learning
RH-BrainFS: Regional Heterogeneous Multimodal Brain Networks Fusion Strategy
Hongting Ye, Yalu Zheng, Yueying Li, Ke Zhang, Youyong Kong, Yonggui Yuan
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add BrainSubGNN and Trans-Bottleneck together versus both replaced by GIN and standard transformer (ablation) | RH-BrainFS · HCP | accuracy: 76.23±3.78 -> 78.63±4.36 | Improved |
| add BrainSubGNN (rooted subgraph GIN) instead of whole-graph GIN (ablation) | RH-BrainFS · HCP | accuracy: 77.22±3.89 -> 78.63±4.36 | Improved |
| add Trans-Bottleneck instead of standard transformer with direct interaction (ablation) | RH-BrainFS · HCP | accuracy: 76.80±2.93 -> 78.63±4.36 | Improved |
| add Trans-Bottleneck instead of standard transformer (AUC) (ablation) | RH-BrainFS · HCP | AUC: 76.38±3.27 -> 78.42±4.38 | Improved |
| number of bottlenecks 2 versus 4 (ablation) | RH-BrainFS · HCP | accuracy: 78.63±4.36 -> 76.26±3.58 | Hurt |
| number of bottlenecks 8 versus 4 (ablation) | RH-BrainFS · Two-site | accuracy: 78.48±1.43 -> 76.71±1.32 | Hurt |
| number of bottlenecks 6 versus 4 (ablation) | RH-BrainFS · Two-site | accuracy: 78.48±1.43 -> 76.35±1.43 | Hurt |
| any number of bottlenecks 2-8 versus baselines (ablation) | RH-BrainFS · HCP and Two-site | accuracy | Improved |
| subgraph sampling hops increased from 1 to 5 (ablation) | RH-BrainFS · HCP | accuracy: 78.63±4.36 -> 73.92±2.94 | Hurt |
| subgraph sampling hops increased from 1 to 2 (ablation) | RH-BrainFS · Two-site | accuracy: 78.48±1.43 -> 73.99±1.32 | Hurt |
| subgraph sampling hops 3 versus 1 (ablation) | RH-BrainFS · HCP | accuracy: 78.63±4.36 -> 77.47±3.31 | Hurt |
| vary adjacency thresholding value 0.02-0.30 (ablation) | RH-BrainFS · HCP | accuracy | No clear effect |
Closest papers by meaning
- Brain Harmony: A Multimodal Foundation Model Unifying Morphology and Function into 1D Tokens (NeurIPS 2025 · similarity 0.88)
- A Local-Ascending-Global Learning Strategy for Brain-Computer Interface (AAAI 2024 · similarity 0.87)
- $\textit{NeuroPath}$: A Neural Pathway Transformer for Joining the Dots of Human Connectomes (NeurIPS 2024 · similarity 0.86)
- BrainOOD: Out-of-distribution Generalizable Brain Network Analysis (ICLR 2025 · similarity 0.85)
- Brain-OF: An Omnifunctional Foundation Model for fMRI, EEG and MEG (NeurIPS 2026 · similarity 0.85)
- Multi-Region Markovian Gaussian Process: An Efficient Method to Discover Directional Communications Across Multiple Brain Regions (ICML 2024 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
