Toward a Well-Calibrated Discrimination via Survival Outcome-Aware Contrastive Learning
This addresses the problem of poor calibration in survival analysis models for clinical applications, though it is incremental as it builds on existing contrastive learning and ranking loss methods.
The paper tackled the trade-off between discrimination and calibration in deep learning for survival analysis by proposing a novel contrastive learning approach with weighted sampling, which improved both discrimination and calibration on multiple clinical datasets, outperforming state-of-the-art models.
Previous deep learning approaches for survival analysis have primarily relied on ranking losses to improve discrimination performance, which often comes at the expense of calibration performance. To address such an issue, we propose a novel contrastive learning approach specifically designed to enhance discrimination \textit{without} sacrificing calibration. Our method employs weighted sampling within a contrastive learning framework, assigning lower penalties to samples with similar survival outcomes. This aligns well with the assumption that patients with similar event times share similar clinical statuses. Consequently, when augmented with the commonly used negative log-likelihood loss, our approach significantly improves discrimination performance without directly manipulating the model outputs, thereby achieving better calibration. Experiments on multiple real-world clinical datasets demonstrate that our method outperforms state-of-the-art deep survival models in both discrimination and calibration. Through comprehensive ablation studies, we further validate the effectiveness of our approach through quantitative and qualitative analyses.