Getting Ready for Generative AI in Academic Medical Centers

This video provides a wide-ranging discussion of data acquisition, analytics, and AI in health care involving Dr. Keith Morse, MD, Professor, Pediatrics, and Medical Director of Clinical Informatics at Stanford Children’s Health, and Dr. Armando Bedoya, MD, Chief Analytics and Medical Informatics Officer at Duke Health.  They had both just spoken at the SAS Innovate 2024 conference about the use of AI, data, and analytics in their organizations.

Both are wary of introducing generative AI too soon, before clinicians and data scientists understand where it is actually useful and how to handle its risks. Morse says that Stanford is waiting while its staff “play around” with generative AI and they learn more at conferences, etc. Bedoya says that Duke holds training sessions where they explain such details as how an AI model was trained so that users can evaluate its value and risks.

On the other hand, according to Morse, large language models allow a much faster applicaiton of natural language processing, which used to require a large development effort for every specialized application.

Bedoya also addresses AI bias in the video, saying that monitoring for bias is “embedded in every single checkpoint that we have.” Morse is concerned with making sure that the pediatric population—often under-represented in medical research—is represented in the data used to develop AI models.

Bedoya described a center at Duke pulling together subject matter experts with technologists and legal and regulatory staff to suppor the responsible use of AI. They determine the risk of each application, how long it is expected to be in use, and other factors.

Morse describes Stanford’s “federated data approach” that distributes AI experts among different departments in order to provide more efficient service in the face of growing requests for AI.

Watch the video for more insights into how analytics are being used in these academic centers and how they are dealing with analytics as their workflows grow increasingly complex.

Learn more about Stanford Children’s Health:

Learn more about Duke Health:

Learn more about SAS:

Listen and subscribe to the Healthcare IT Today Interviews Podcast to hear all the latest insights from experts in healthcare IT.

And for an exclusive look at our top stories, subscribe to our newsletter and YouTube.

Tell us what you think. Contact us here or on Twitter at @hcitoday. And if you’re interested in advertising with us, check out our various advertising packages and request our Media Kit.