Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jul 3, 2026

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

Temporally Phenotyping GLP-1RA Case Reports with Large Language Models: A Textual Time Series Corpus and Risk

Sayantan Kumar1, Jeremy C Weiss1

  • 1National Library of Medicine, Bethesda, Maryland, USA.

Arxiv
|July 2, 2026
PubMed
Summary

We developed a method to extract timelines from type 2 diabetes case reports using large language models (LLMs). This approach accurately captures clinical events and timings, aiding longitudinal modeling and demonstrating potential benefits of glucagon-like peptide-1 receptor agonists.

Related Concept Videos

Introduction to Language of Pathophysiology ll01:17

Introduction to Language of Pathophysiology ll

This lesson explores key terms that describe how diseases progress, their outcomes, and their distribution in populations.Diagnostic tests identify diseases and monitor treatment. These include blood and urine tests, biopsies, imaging (X-ray, MRI), and detection of infectious agents.Remission is a reduction or disappearance of symptoms.Exacerbation refers to the worsening of symptoms, such as increased wheezing during an asthma attack.A precipitating factor triggers an acute episode, while a...

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Text Knows What, Tables Know When: Clinical Timeline Reconstruction via Retrieval-Augmented Multimodal Alignment.

ArXiv·2026
Same author

Hierarchical Barycentric Multimodal Representation Learning for Medical Image Analysis.

medRxiv : the preprint server for health sciences·2026
Same author

Temporally Phenotyping GLP-1RA Case Reports with Large Language Models: A Textual Time Series Corpus and Risk Modeling.

medRxiv : the preprint server for health sciences·2026
Same author

Forecasting Clinical Risk from Textual Time Series: Structuring Narratives for Temporal AI in Healthcare.

medRxiv : the preprint server for health sciences·2025
Same author

The Impact of Medication Non-adherence on Adverse Outcomes: Evidence from Schizophrenia Patients via Survival Analysis.

Proceedings of machine learning research·2025
Same author

Multimodal Variational Autoencoder: A Barycentric View.

Proceedings of the ... AAAI Conference on Artificial Intelligence. AAAI Conference on Artificial Intelligence·2025

Area of Science:

  • Medical Informatics
  • Clinical Data Science
  • Pharmacology

Background:

  • Type 2 diabetes case reports contain complex clinical timelines difficult for longitudinal modeling.
  • Extracting and standardizing temporal event data from clinical narratives is a significant challenge.

Purpose of the Study:

  • To develop a textual time-series corpus from PubMed case reports on glucagon-like peptide-1 receptor agonists (GLP-1 RAs).
  • To evaluate automated timeline extraction using large language models (LLMs) against expert-annotated timelines.
  • To assess LLM performance in recovering clinical events and their timings for longitudinal analysis.

Main Methods:

  • Created a corpus of 136 PubMed Open Access single-patient case reports involving GLP-1 RAs.
  • Associated clinical events with probable reference times within the case reports.

Related Experiment Videos

Last Updated: Jul 3, 2026

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

  • Evaluated automated LLM timeline extraction (e.g., GPT5) against gold-standard timelines annotated by clinical experts.
  • Assessed event coverage and temporal sequencing accuracy of LLM-extracted timelines.

    • Main Results:

    • The best-performing LLM (GPT5) achieved high event coverage (0.871) and reliable temporal sequencing (0.843).
    • LLM extraction accurately captured timings for symptoms, diagnoses, treatments, lab tests, and outcomes.
    • Downstream analysis showed GLP-1 RA users had a lower risk of respiratory sequelae (HR=0.259, p<0.05).

    Conclusions:

    • Automated LLM timeline extraction is effective for standardizing complex clinical data from type 2 diabetes case reports.
    • This methodology facilitates longitudinal modeling and analysis of treatment effects.
    • Findings support previous reports of improved respiratory outcomes associated with GLP-1 RA use.