Toward a Well-Calibrated Discrimination via Survival Outcome-Aware Contrastive Learning

TL;DR

This paper introduces a contrastive learning method for survival analysis that improves discrimination without harming calibration, using weighted sampling based on survival outcome similarity.

Contribution

The proposed approach uniquely combines contrastive learning with survival analysis to enhance discrimination while maintaining calibration, outperforming existing models.

Findings

Outperforms state-of-the-art models in discrimination and calibration

Effective in real-world clinical datasets

Validated through extensive ablation studies

Abstract

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.