Harnessing large language models in laboratory medicine

Dr. Joe El-Khoury, Director of Clinical Chemistry at Yale New Haven Health, discusses the potential of large language models (LLMs) like ChatGPT in laboratory medicine, particularly as tools for providing clinical lab consults. Recent studies show that LLMs can provide answers that are preferred over those given by doctors in 76% of cases. However, ensuring the accuracy and validity of these models is a significant challenge. Continuous monitoring and validation are essential, especially as these models are frequently updated with new training data, making it complex to ensure consistent, accurate performance in clinical settings.

This interview was recorded for #CLINICAL24 2024.