Scaling AI for Real-World Radiology Workflows
 16 snips 
 Mar 25, 2025  Dr. Pranav Rajpurkar, a pioneer in AI at Harvard, and Dr. Michael Moritz, an abdominal radiologist from St. Louis University, dive into the future of radiology. They discuss how AI can revolutionize disease detection and improve clinical workflows, while also sharing their entrepreneurial journeys in this tech-driven field. The conversation touches on the importance of collaboration between radiologists and AI experts to achieve ambitious goals, like minimizing errors in diagnoses, and emphasizes the need for adaptability in the face of rapid technological advancements. 
 AI Snips 
 Chapters 
 Transcript 
 Episode notes 
One Model for Many Diseases
- Dr. Rajpurkar realized the mismatch between AI development and radiology workflow.
 - Building one model to detect many diseases is more effective than many single-disease models.
 
Generalist AI for Medical Imaging
- Training generalist AI models with medical data is the key.
 - A good fracture detection model comes from training a model on diverse medical images and tasks.
 
Medical AI Bootcamp Collaboration
- Dr. Moritz met Dr. Rajpurkar in his medical AI bootcamp.
 - The bootcamp brought together radiologists and computer science students for collaboration.
 
