Research map / Vision-language and multimodal / Cross-modal learning
GTP-4o: Modality-prompted Heterogeneous Graph Learning for Omni-modal Biomedical Representation
Chenxin Li, Xinyu Liu, Cheng Wang, Yifan Liu, Weihao Yu, Jing Shao, Yixuan 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 graph-prompted completion instead of zero-filling missing modality (ablation) | GTP-4o · TCGA-GBMLGG | Glioma grading AUC: 0.9087 -> 0.9389 | Improved |
| add graph-prompted completion instead of dropping missing modality (ablation) | GTP-4o · TCGA-GBMLGG | Glioma grading AUC: 0.9288 -> 0.9389 | Improved |
| add heterogeneous embedding instead of homogeneous graph (ablation) | GTP-4o · TCGA-GBMLGG | Glioma grading AUC: 0.9232 -> 0.9389 | Improved |
| add heterogeneous relations on edges (ablation) | GTP-4o · TCGA-GBMLGG | Survival C-Index: 0.8201 -> 0.8416 | Improved |
| add prompt bank (ablation) | GTP-4o completion · TCGA-GBMLGG | Survival C-Index: 0.8280 -> 0.8416 | Improved |
| add knowledge-guided hierarchical aggregation instead of plain mean (ablation) | GTP-4o · TCGA-GBMLGG | Survival C-Index: 0.8311 -> 0.8416 | Improved |
| use knowledge-derived meta-paths instead of random meta-paths (ablation) | GTP-4o aggregation · TCGA-GBMLGG | Glioma grading AUC: 0.9350 -> 0.9389 | Improved |
| add cell graph and text modalities to genomics and images | GTP-4o · TCGA-GBMLGG | Glioma grading AUC: 0.9256 -> 0.9336 -> 0.9389 | Improved |
| use multimodal (G+I) instead of unimodal inputs | baselines and GTP-4o · TCGA-GBMLGG | Glioma grading AUC: GTP-4o G+I 0.9256 vs HEAT image-only 0.9289 | Mixed |
| GTP-4o versus TransFusion with four modalities | GTP-4o · TCGA-KIRC | Survival C-Index: 0.7289 -> 0.7336 | Improved |
Closest papers by meaning
- Modalities Contribute Unequally: Enhancing Medical Multi-modal Learning through Adaptive Modality Token Re-balancing (ICML 2025 · similarity 0.87)
- Flex-MoE: Modeling Arbitrary Modality Combination via the Flexible Mixture-of-Experts (NeurIPS 2024 · similarity 0.86)
- Knowledge Bridger: Towards Training-Free Missing Modality Completion (CVPR 2025 · similarity 0.86)
- Jointly Modeling Inter- & Intra-Modality Dependencies for Multi-modal Learning (NeurIPS 2024 · similarity 0.86)
- Multi-modal Vision Pre-training for Medical Image Analysis (CVPR 2025 · similarity 0.86)
- Gradient-Guided Modality Decoupling for Missing-Modality Robustness (AAAI 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.
