One Loss to Rule Them All: Marked Time-to-Event for Structured EHR Foundation Models

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Jing, Zilin, Jeanselme, Vincent, Kobayashi, Yuta, Lee, Simon A., Pang, Chao, Kashyap, Aparajita, Li, Yanwei, Jiang, Xinzhuo, Joshi, Shalmali
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918317170622464
author Jing, Zilin
Jeanselme, Vincent
Kobayashi, Yuta
Lee, Simon A.
Pang, Chao
Kashyap, Aparajita
Li, Yanwei
Jiang, Xinzhuo
Joshi, Shalmali
author_facet Jing, Zilin
Jeanselme, Vincent
Kobayashi, Yuta
Lee, Simon A.
Pang, Chao
Kashyap, Aparajita
Li, Yanwei
Jiang, Xinzhuo
Joshi, Shalmali
contents Clinical events captured in Electronic Health Records (EHR) are irregularly sampled and may consist of a mixture of discrete events and numerical measurements, such as laboratory values or treatment dosages. The sequential nature of EHR, analogous to natural language, has motivated the use of next-token prediction to train prior EHR Foundation Models (FMs) over events. However, this training fails to capture the full structure of EHR. We propose ORA, a marked time-to-event pretraining objective that jointly models event timing and associated measurements. Across multiple datasets, downstream tasks, and model architectures, this objective consistently yields more generalizable representations than next-token prediction and pretraining losses that ignore continuous measurements. Importantly, the proposed objective yields improvements beyond traditional classification evaluation, including better regression and time-to-event prediction. Beyond introducing a new family of FMs, our results suggest a broader takeaway: pretraining objectives that account for EHR structure are critical for expanding downstream capabilities and generalizability
format Preprint
id arxiv_https___arxiv_org_abs_2602_00541
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle One Loss to Rule Them All: Marked Time-to-Event for Structured EHR Foundation Models
Jing, Zilin
Jeanselme, Vincent
Kobayashi, Yuta
Lee, Simon A.
Pang, Chao
Kashyap, Aparajita
Li, Yanwei
Jiang, Xinzhuo
Joshi, Shalmali
Machine Learning
Clinical events captured in Electronic Health Records (EHR) are irregularly sampled and may consist of a mixture of discrete events and numerical measurements, such as laboratory values or treatment dosages. The sequential nature of EHR, analogous to natural language, has motivated the use of next-token prediction to train prior EHR Foundation Models (FMs) over events. However, this training fails to capture the full structure of EHR. We propose ORA, a marked time-to-event pretraining objective that jointly models event timing and associated measurements. Across multiple datasets, downstream tasks, and model architectures, this objective consistently yields more generalizable representations than next-token prediction and pretraining losses that ignore continuous measurements. Importantly, the proposed objective yields improvements beyond traditional classification evaluation, including better regression and time-to-event prediction. Beyond introducing a new family of FMs, our results suggest a broader takeaway: pretraining objectives that account for EHR structure are critical for expanding downstream capabilities and generalizability
title One Loss to Rule Them All: Marked Time-to-Event for Structured EHR Foundation Models
topic Machine Learning
url https://arxiv.org/abs/2602.00541