AI for healthcare: Scaling Access and Quality of Care for Everyone
summary
Curai deploys a mobile primary-care app with hundreds of thousands of users that combines NLP, vision and automatic diagnosis by mixing legacy expert systems with deep learning; a clinical case simulator generates training data from the expert knowledge base, and prototypical clustering networks enable few-shot extension. The stack models unknowns with open-set diagnosis so automation falls back to a physician in the loop. Claims: <$0.20 of today’s cost, extensible models; five accepted papers; context: 50% lack basic care, 30% US underinsured, 120k physician shortfall by 2030, 60% single-doctor accuracy vs 85% consensus.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 0:56The problem
- 4:10What were people doing
- 5:06AI for healthcare
- 6:57Our mission
- 7:44Our research
- 11:50Functionality of an AI system
- 13:16Basic building blocks
- 15:52Are AI systems ready
- 17:26Extensibility
- 21:38Knowing what you dont know
- 23:36extensible
related talks
29mAI for COVID-19: An online virtual care approach
38mAI Impact in Healthcare
47mThe Future of Healthcare and AI
36mAI in healthcare - what is hype and what is real?
47mAI for making decisions in clinical medicine
28mAI in healthcare - Deep learning for precision psychiatry
37mAI in medicine: where do we stand?
48mAI for Healthcare