The AI PHI Affinity Group hosted Adam Yala — Assistant Professor of Computational Precision Health and EECS at UC Berkeley and UCSF — for a December 2022 talk titled “Seeing into the future: Machine learning methods for personalized screening.” Yala’s research sits at the intersection of risk prediction, screening-policy design, and privacy-preserving data sharing — three threads that bear directly on AI PHI’s mission of reducing cancer burden through earlier and smarter detection.

The talk centered on the tension between early-detection benefit and overscreening harm. Risk models guide screening decisions for millions of patients each year, but most current models are calibrated to populations and imaging conditions that don’t generalize cleanly to the multiethnic communities AI PHI serves. Yala walked through AI approaches to imaging-based risk assessment and to designing screening policies that remain robust under data-generation biases — the kind of mismatch that can quietly degrade model performance once deployed in a new clinic or for a different patient subgroup.

Several of Yala’s models have moved into prospective clinical trials, and his work has been covered in the Washington Post, New York Times, Boston Globe, and Wired. He holds BS, MEng, and PhD degrees in Computer Science from MIT, and was affiliated with the MIT Jameel Clinic and MIT CSAIL before his current appointment.

A recording of the talk is available on the AI PHI YouTube channel ↗. The Affinity Group meets the first Friday of each month at 9:00 AM Hawaiʻi time — see the events page for upcoming speakers.