Score: 1

TIMER: Temporal Instruction Modeling and Evaluation for Longitudinal Clinical Records

Published: March 6, 2025 | arXiv ID: 2503.04176v1

By: Hejie Cui , Alyssa Unell , Bowen Chen and more

BigTech Affiliations: Stanford University

Potential Business Impact:

Helps computers understand patient health history over time.

Business Areas:
Electronic Health Record (EHR) Health Care

Large language models (LLMs) have emerged as promising tools for assisting in medical tasks, yet processing Electronic Health Records (EHRs) presents unique challenges due to their longitudinal nature. While LLMs' capabilities to perform medical tasks continue to improve, their ability to reason over temporal dependencies across multiple patient visits and time frames remains unexplored. We introduce TIMER (Temporal Instruction Modeling and Evaluation for Longitudinal Clinical Records), a framework that incorporate instruction-response pairs grounding to different parts of a patient's record as a critical dimension in both instruction evaluation and tuning for longitudinal clinical records. We develop TIMER-Bench, the first time-aware benchmark that evaluates temporal reasoning capabilities over longitudinal EHRs, as well as TIMER-Instruct, an instruction-tuning methodology for LLMs to learn reasoning over time. We demonstrate that models fine-tuned with TIMER-Instruct improve performance by 7.3% on human-generated benchmarks and 9.2% on TIMER-Bench, indicating that temporal instruction-tuning improves model performance for reasoning over EHR.

Country of Origin
🇺🇸 United States

Page Count
19 pages

Category
Computer Science:
Artificial Intelligence