Well Beyond Medicine: The Nemours Children's Health Podcast · Nemours Children's Health

Ep. 169: Exploring Mayo Clinic's AI Evolution with Micky Tripathi, PhD, MPP

·30 min·1 clip

Tripathi says AI could move medicine from evidence-based to “medicine-based evidence.”

1. Well Beyond Medicine: The Nemours Children's Health Podcast episode focuses on Mayo Clinic's AI evolution with Micky Tripathi, PhD, MPP, at Health in Las Vegas. 2. Carol Vassar hosts Tripathi, Mayo Clinic’s Chief Artificial Intelligence Implementation Officer, whose role centers on turning AI into front-line clinical practice. 3. The episode asks why Mayo Clinic is pushing AI implementation now and how that changes patient care across the enterprise. 4. Tripathi says he moved from four years in federal technology policy and chief AI work into Mayo because he wanted to implement technology at the point of care. 5. He says Mayo Clinic’s patient-first culture and resources made it a natural place to focus on AI that reaches physicians quickly. 6. He says he was surprised by the amount of innovation already underway when he arrived at Mayo. 7. Tripathi gives a concrete snapshot: 97 AI solutions are in production, more than 270 are in the pipeline, and most are predictive AI with some generative AI. 8. He says all of those systems have human oversight, which he describes as human-in-the-loop protection. 9. He says Mayo also works with EPIC, third-party tools, the Mayo Clinic platform, and Mayo innovators, so the AI stack is not limited to one source. 10. He describes Mayo-built multimodal foundation models for genetic encoding and digital pathology, including a model trained on 16 million digitized glass slides. 11. He says the pathology work uses Mayo’s longitudinal record to look back at pre-diagnosis slides and ask whether AI can detect early signs of cancer that human review missed. 12. He adds that Mayo is rolling out radiology foundation models that look for early pancreatic cancer and help with chest X-ray review. 13. Tripathi says one chest X-ray model may save a radiologist 30 to 45 seconds per image, which matters when a radiologist has 125 chest X-rays in a morning. 14. He explains that Mayo is collaborating through coalitions such as the Coalition for Health AI and Health AI Partnership to standardize validation frameworks. 15. He says the Mayo Clinic platform, run by John Halamka, was built four years ago to share de-identified data with providers like Mercy Health, Israeli hospitals, Singapore, and Brazil. 16. He argues that AI implementation has slowed because privacy, IRB, and technology assurance bodies face a surge of complex cases that do not fit neatly into one review lane. 17. He says the governance challenge is not only bias, privacy, security, and hallucinations, but also the opportunity cost of not deploying helpful tools quickly. 18. He uses rheumatoid arthritis and methotrexate to show why early genetic prediction could matter when 30 to 50 percent of patients do not respond to the standard treatment. 19. The interview is direct, technical, and conversational, with Carol Vassar pressing for concrete examples and Tripathi answering in detailed, practical terms. 20. Listeners interested in clinical AI, precision medicine, or hospital implementation will get the most from this episode; listeners wanting light health news may skip it.

As heard by us

A grounded look at AI, frontline value, and the cost of waiting too long.

The episode keeps its focus on a practical question: how Mickey Tripathi sees AI moving from policy talk into a large health system that has to use it carefully, protect patients, and still make care better.

Read the full review in PlayNext →

Why you'd press play

For a look at how Mayo Clinic is trying to bring AI into real clinical workflows, start here.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Well Beyond Medicine: The Nemours Children's Health Podcast