ncibtep@nih.gov

Bioinformatics Training and Education Program

Leveraging Large Language Models for Diagnostic Safety and Quality Improvement

Leveraging Large Language Models for Diagnostic Safety and Quality Improvement

 When: Jul. 10th, 2025 2:00 pm - 3:00 pm

Learning Level: Any

To Know

Where:
Online Webinar
Organizer:
NIH Text Mining and Natural Language Processing SIG
Presented By:
Andrew Taylor (UVA)

About this Class

Diagnostic errors remain a leading cause of preventable harm in healthcare, particularly in high-stakes, time-pressured environments like the emergency department. Large Language Models (LLMs) offer promising new capabilities for transforming diagnostic safety surveillance and clinical decision support. In this talk, Dr. Taylor will explore emerging applications of LLMs to improve diagnostic performance, with a focus on real-time chart review, trigger-based systems, and the augmentation of quality improvement processes. He will share recent research evaluating the feasibility, accuracy, and implementation challenges of these models, as well as lessons learned from deploying them in clinical and operational workflows. We’ll also discuss the broader implications for trust, oversight, and future regulatory frameworks as we bring these tools closer to the bedside.