Panel Discusses AI Applications Across Healthcare Continuum

A panel at the Imagination in Action event in April examined how artificial intelligence fits into the continuum of care, defined as the full range of healthcare services a person may need over time. Emily Capodilupo of Whoop moderated the discussion, which included Eric Rosenthal of the MGB NeuroAI Center, Gokul Radhakrishnan of Eli Lilly, Brad Reimer of Sanford Health, and Sufian Chowdhury of Kinetik.

Rosenthal said models now used in imaging sometimes fail to match expectations and lack the data practices needed for trustworthy systems. Radhakrishnan stated the FDA seeks transparent validation processes rather than black-box outputs. Reimer noted that many workflows remain tied to electronic medical record systems introduced two decades ago.

Chowdhury said reliance on subjective data for clinical decisions could lead to poor outputs without stronger protocols. Reimer added that governance structures are still developing and must address rapid overlap among new tools. Chowdhury also said government agencies lack sufficient AI expertise and called for expanded public-private partnerships.

Rosenthal described scribe-style AI tools that reduce time physicians spend on documentation during patient visits. Radhakrishnan cited a Deloitte study showing reduced documentation time per physician and noted that newer ambient systems differ from earlier wearable devices. Rosenthal said clinicians should verify AI outputs rather than accept them without review.