Research map / Graph learning / GNN architectures
MAPI-GNN: Multi-Activation Plane Interaction Graph Neural Network for Multimodal Medical Diagnosis
Ziwei Qin, Xuhui Song, Deqing Huang, Na Qin, Jun Li
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 multi-dimensional feature discriminator (ablation) | MAPI-GNN · PI-CAI (csPCa, 440 cases, 5-fold CV) | ACC: 0.8500 -> 0.9432 | Improved |
| add multi-activation graph construction (ablation) | MAPI-GNN · PI-CAI (csPCa, 440 cases, 5-fold CV) | ACC: 0.8205 -> 0.9432 | Improved |
| add hierarchical feature dynamic association network (ablation) | MAPI-GNN · PI-CAI (csPCa, 440 cases, 5-fold CV) | ACC: 0.8364 -> 0.9432 | Improved |
| add MDFD (ablation) | MAPI-GNN · CHD (CCTA + clinical, 974 cases) | ACC: 0.8333 -> 0.9027 | Improved |
| add MAGCS (ablation) | MAPI-GNN · CHD (CCTA + clinical, 974 cases) | ACC: 0.8673 -> 0.9027 ACC; AUC 0.9538 -> 0.9206 | Mixed |
| add HFDAN (ablation) | MAPI-GNN · CHD (CCTA + clinical, 974 cases) | ACC: 0.8584 -> 0.9027 ACC; AUC 0.9306 -> 0.9206 | Mixed |
| increase semantic dimensions M from 8 to 24 (ablation) | MAPI-GNN · PI-CAI (csPCa, 440 cases, 5-fold CV) | ACC: 0.9136 -> 0.9432 | Improved |
| vary neighbours k from 5 to 3 or 10 (ablation) | MAPI-GNN · PI-CAI (csPCa, 440 cases, 5-fold CV) | ACC: 0.9432 (k=5) vs 0.9091 (k=3), 0.9205 (k=10) | Hurt |
| vary activated feature proportion (ablation) | MAPI-GNN · PI-CAI (csPCa, 440 cases, 5-fold CV) | ACC: 0.9430 (3%), 0.9432 (5%), 0.9424 (8%), 0.9409 (10%) | No clear effect |
| use zeroing-out instead of halving or set-to-1 perturbation (ablation) | MAPI-GNN · PI-CAI (csPCa, 440 cases, 5-fold CV) | ACC: 0.9409 (set to 1), 0.9421 (halving) -> 0.9432 | Improved |
| GNN-based method over CNN/ViT baselines | HGM2R · PI-CAI (csPCa, 440 cases, 5-fold CV) | ACC: 0.8522 (MSC) -> 0.9242 (HGM2R) | Improved |
Closest papers by meaning
- Graph Neural Networks in Network Neuroscience (TPAMI 2022 · similarity 0.84)
- BrainMAP: Learning Multiple Activation Pathways in Brain Networks (AAAI 2025 · similarity 0.83)
- Visual and Domain Knowledge for Professional-level Graph-of-Thought Medical Reasoning (ICML 2025 · similarity 0.82)
- DisenGCD: A Meta Multigraph-assisted Disentangled Graph Learning Framework for Cognitive Diagnosis (NeurIPS 2024 · similarity 0.82)
- BrainOOD: Out-of-distribution Generalizable Brain Network Analysis (ICLR 2025 · similarity 0.82)
- MKG-FENN: A Multimodal Knowledge Graph Fused End-to-End Neural Network for Accurate Drug–Drug Interaction Prediction (AAAI 2024 · similarity 0.82)
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.
