Counterfactual Understanding via Retrieval-aware Multimodal Modeling for Time-to-Event Survival Prediction
Ha-Anh Hoang Nguyen, Tri-Duc Phan Le, Duc-Hoang Pham, Huy-Son Nguyen, Cam-Van Thi Nguyen, Duc-Trong Le, Hoang-Quynh Le · 2026
Read Counterfactual Understanding via Retrieval-aware Multimodal Modeling for Time-to-Event Survival Prediction by Ha-Anh Hoang Nguyen, Tri-Duc Phan Le, Duc-Hoang Pham, Huy-Son Nguyen, Cam-Van Thi Nguyen, Duc-Trong Le, Hoang-Quynh Le (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-19987.txt.
Description
Predicting how long a patient survives under different treatments is hard when patients differ, data are censored, and the question is what-if: what would happen under another treatment? A 2026 arXiv paper by Ha-Anh Hoang Nguyen and six co-authors proposes CURE, a framework for what-if survival prediction from multi-source data.
CURE fuses clinical, lab-based, demographic, and multi-omics information through cross-attention, refining complex omics signals with a mixture-of-experts design that emphasizes the most informative components. From that picture it retrieves patient-specific latent subgroups capturing baseline survival patterns and treatment-dependent changes. Tests on two cancer data sets show CURE steadily beating strong baselines on standard survival metrics. The work offers a foundation for future treatment recommendation models, with code and resources public.
Catalog record
| CREATOR | Ha-Anh Hoang Nguyen, Tri-Duc Phan Le, Duc-Hoang Pham, Huy-Son Nguyen, Cam-Van Thi Nguyen, Duc-Trong Le, Hoang-Quynh Le |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 34 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-19987