ACG 2025: Large Language Models for Eosinophilic Esophagitis Detection

We are presenting a poster at the American College of Gastroenterology (ACG) Annual Scientific Meeting, which will be held October 24–29, 2025, in Phoenix, AZ.

Poster Presentation

Title: Large Language Models Combined with a Single Diagnostic Code Detect Eosinophilic Esophagitis in Electronic Health Records with High Accuracy

Authors:

• Corey James Ketchem
• Ugurcan Vurgun
• Andrew Wang
• Sarah Thomas
• Alexis Batugo
• Gary W. Falk
• Kristel L. Lynch
• Evan S. Dellon
• Danica L. Mowery
• James D. Lewis

Publication: American Journal of Gastroenterology. 2025; Abstract P2739.

Research Focus: This study demonstrates the effectiveness of large language models (LLMs) combined with minimal diagnostic coding for identifying eosinophilic esophagitis (EoE) cases in electronic health records.

Key Innovation: The method leverages natural language processing to extract meaningful clinical information from EHR text, reducing the need for extensive manual chart review while maintaining high detection accuracy.

Impact: This work has implications for:

• Automated case identification in retrospective studies
• Population health surveillance for rare diseases
• Reducing manual chart review burden in clinical research
• Scalable EHR-based phenotyping strategies

The corresponding manuscript has also been submitted to Clinical Gastroenterology and Hepatology for peer review.

If you’re interested in this research or potential collaborations in AI applications for gastroenterology, please don’t hesitate to reach out.