Building the EHR Foundation Model via Next Event Prediction
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914066933481472 |
|---|---|
| author | Chen, Zekai Pekis, Arda Brown, Kevin |
| author_facet | Chen, Zekai Pekis, Arda Brown, Kevin |
| contents | Electronic Health Records (EHRs) contain rich temporal dynamics that conventional encoding approaches fail to adequately capture. While Large Language Models (LLMs) show promise for EHR modeling, they struggle to reason about sequential clinical events and temporal dependencies. We propose Next Event Prediction (NEP), a framework that enhances LLMs' temporal reasoning through autoregressive fine-tuning on clinical event sequences. By reformulating EHRs as timestamped event chains and predicting future medical events, NEP explicitly models disease progression patterns and causal relationships. Extensive evaluations across oncology survival prediction and clinical diagnosis tasks demonstrate NEP's superiority, outperforming specialized EHR models by 4.6% AUROC and general-purpose LLMs by 7.2% C-index in temporal reasoning tasks. Our analyses reveal dual benefits: state-of-the-art prediction accuracy combined with clinically interpretable attention patterns that align with known disease pathways. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_25591 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Building the EHR Foundation Model via Next Event Prediction Chen, Zekai Pekis, Arda Brown, Kevin Artificial Intelligence Computation and Language Other Quantitative Biology Electronic Health Records (EHRs) contain rich temporal dynamics that conventional encoding approaches fail to adequately capture. While Large Language Models (LLMs) show promise for EHR modeling, they struggle to reason about sequential clinical events and temporal dependencies. We propose Next Event Prediction (NEP), a framework that enhances LLMs' temporal reasoning through autoregressive fine-tuning on clinical event sequences. By reformulating EHRs as timestamped event chains and predicting future medical events, NEP explicitly models disease progression patterns and causal relationships. Extensive evaluations across oncology survival prediction and clinical diagnosis tasks demonstrate NEP's superiority, outperforming specialized EHR models by 4.6% AUROC and general-purpose LLMs by 7.2% C-index in temporal reasoning tasks. Our analyses reveal dual benefits: state-of-the-art prediction accuracy combined with clinically interpretable attention patterns that align with known disease pathways. |
| title | Building the EHR Foundation Model via Next Event Prediction |
| topic | Artificial Intelligence Computation and Language Other Quantitative Biology |
| url | https://arxiv.org/abs/2509.25591 |