Research map / Vision-language and multimodal / Visual question answering
Multimodal Co-Attention Transformer for Survival Prediction in Gigapixel Whole Slide Images
Richard J. Chen, Ming Y. Lu, Wei-Hung Weng, Tiffany Y. Chen, Drew F.K. Williamson, Trevor Manz, Maha Shady, Faisal Mahmood
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 genomic features via co-attention instead of WSI only (MCAT vs Attention MIL WSI Only) | Attention MIL · TCGA 5 cancers overall | c-Index: 0.614 -> 0.653 | Improved |
| use MCAT instead of DeepAttnMISL WSI only | DeepAttnMISL · TCGA 5 cancers overall | c-Index: 0.581 -> 0.653 | Improved |
| use co-attention fusion instead of concatenation late fusion | Attention MIL (Concat) · TCGA 5 cancers overall | c-Index: 0.630 -> 0.653 | Improved |
| use co-attention fusion instead of bilinear pooling late fusion | Attention MIL (Bilinear Pooling) · TCGA 5 cancers overall | c-Index: 0.611 -> 0.653 | Improved |
| use MCAT instead of DeepAttnMISL concat fusion | DeepAttnMISL (Concat) · TCGA BRCA | c-Index: 0.545 ± 0.071 -> 0.580 ± 0.069 | Improved |
| add genomic concat to Attention MIL (WSI only) in UCEC | Attention MIL · TCGA UCEC | c-Index: 0.625 ± 0.057 -> 0.614 ± 0.052 | Hurt |
| add genomic concat to Deep Sets WSI only | Deep Sets · TCGA 5 cancers overall | c-Index: 0.499 -> 0.629 | Improved |
| add Transformer layers before global attention pooling (both TH and TG) (ablation) | MCAT · TCGA overall c-Index | c-Index | Improved |
Closest papers by meaning
- Multimodal Prototyping for cancer survival prediction (ICML 2024 · similarity 0.90)
- Interpretable Vision-Language Survival Analysis with Ordinal Inductive Bias for Computational Pathology (ICLR 2025 · similarity 0.90)
- Modeling Dense Multimodal Interactions Between Biological Pathways and Histology for Survival Prediction (CVPR 2024 · similarity 0.89)
- CA-MLIF: Cross-Attention and Multimodal Low-Rank Interaction Fusion Framework for Tumor Prognostic Prediction (AAAI 2025 · similarity 0.88)
- Robust Multimodal Survival Prediction with Conditional Latent Differentiation Variational AutoEncoder (CVPR 2025 · similarity 0.87)
- Context Matters: Query-aware Dynamic Long Sequence Modeling of Gigapixel Images (ICML 2025 · similarity 0.87)
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.
