Event-Based Contrastive Learning for Medical Time Series

Hyewon Jeong, Nassim Oufattole, Aparna Balagopalan, Matthew B.A. McDermott, Payal Chandak, Marzyeh Ghassemi, Collin Stultz

Keywords: Contrastive-Learning, Time-Series, Healthcare, EHR

TL;DR: We propose a novel contrastive-learning pretraining method for improving EHR time-series representation learning around clinically significant events.

Abstract: In clinical practice, one often needs to identify whether a patient is at high risk of adverse outcomes after some key medical event; e.g., the short-term risk of death after an admission for heart failure. This task, however, remains challenging due to the complexity, variability, and heterogeneity of longitudinal medical data, especially for individuals suffering from chronic diseases like heart failure. In this paper, we introduce Event-Based Contrastive Learning (EBCL) - a method for learning embeddings of heterogeneous patient data that preserves temporal information before and after key index events. We demonstrate that EBCL produces models that yield better fine-tuning performance on critical downstream tasks including 30-day readmission, 1-year mortality, and 1-week length of stay relative to other representation learning methods that do not exploit temporal information surrounding key medical events.

Track: Extended Abstract Track

Submission Number: 24