Radiology AI Podcast | RSNA Episode 61: Radiology: AI RSNA2025 Fireside Chat LIVE
Feb 13, 2026
Linda Moy, MD, radiologist and AI governance lead at NYU; Woojin Kim, MD, MSK radiologist and AI strategy officer; Nina Kottler, MD, MS, applied-math–trained clinical AI implementer; Charles Kahn, Jr., MD, MS, radiology AI journal editor. They debate clinical deployment of foundation models, draft reporting and workflow integration, human-AI collaboration risks and safeguards, multimodality tools, adoption speed, regulation, and governance.
AI Snips
Chapters
Transcript
Episode notes
Foundation Models Are In Clinical Pilots
- Foundation models and automated draft reporting are already being piloted clinically and show meaningful efficiency gains.
- Expect rapid adoption and increasing scrutiny as radiologists begin to demand higher accuracy over time.
Deploy Drafting With Human Oversight
- Deploy vision-language drafting only with governance and monitoring to preserve quality while improving capacity.
- Combine draft reports with radiologist review to decrease fatigue and increase focused image interpretation.
Human–AI Interaction Is The Critical Frontier
- Human–AI interaction is the central problem: the outcome equals the radiologist plus the AI, not the AI alone.
- We must engineer workflows and interfaces so human and model produce an outsized combined effect.