Tabular Foundation Models for Clinical Survival Analysis via Survival-Aware Adaptation
Minh-Khoi Pham, Luca Cotugno, Alina Sirbu, Tai Tan Mai, Martin Crane, Marija Bezbradica
Why It Matters
What makes this one worth your time
This work addresses a critical gap in survival analysis by leveraging advanced machine learning techniques, potentially improving clinical decision-making in critical care settings.
A novel adaptation of tabular foundation models enhances clinical survival prediction accuracy.
Summary
The paper presents a method for adapting tabular foundation models to clinical survival analysis by training a survival-aware head on top of pretrained representations, demonstrating its effectiveness on public survival benchmarks and large-scale ICU cohorts.
Key contributions
- Proposes a lightweight adaptation approach for tabular foundation models in survival analysis.
- Evaluates the method on diverse public survival benchmarks and large-scale ICU datasets.
- Demonstrates competitive performance against established baselines in clinical settings.
Notable insights
- The use of a multi-task logistic regression head allows for effective modeling of right-censored outcomes, which is often challenging in survival analysis.
- The results indicate that transfer learning from tabular foundation models can yield significant performance improvements over traditional methods.
Possible limitations
- Not stated in the abstract.
Abstract
arXiv:2606.12006v1 Announce Type: cross Abstract: Predicting time-to-event outcomes such as mortality is a fundamental task in clinical decision-making, commonly addressed through survival analysis. While classical statistical and deep learning approaches have been widely studied, they typically require task-specific training and sufficient labeled data. Recent advances in tabular foundation models offer a new paradigm by learning general-purpose representations for structured data. However, their applicability to censored time-to-event prediction in clinical settings remains underexplored, as typical applications are restricted to discrete classification rather than survival analysis tasks. In this work, we propose a lightweight adaptation approach for applying tabular foundation models to clinical survival analysis by directly training a survival-aware head on top of the pretrained representations. We study representative architectures, including TabPFN, TabDPT, and TabICL, and adapt them using a multi-task logistic regression (MTLR) head to model right-censored time-to-event outcomes. We evaluate this approach on a diverse set of public survival benchmarks and two large-scale ICU cohorts, MIMIC-IV and eICU. Our results show that this transfer learning approach achieves competitive or superior performance compared to strong baselines. On MIMIC-IV, TabDPT-FT-MTLR reaches a C-index of 0.856, corresponding to a relative improvement of +1.4% over the best non-FM baseline (DeepSurv, 0.844) and +6.7% over the best zero-shot model (0.802). On eICU, TabICL-FT-MTLR achieves 0.797, yielding gains of +1.7% (DeepSurv, 0.784) and +6.4% (0.749), respectively. These findings highlight the importance of combining pretrained tabular representations with survival-aware objectives and suggest that tabular foundation models provide a practical and effective alternative for clinical survival prediction.