AI Happy Hour | AIMI Symposium Q&A

This live panel discussion is part of the AI Happy Hour series, brought to you by Stanford AIMI and friends. We cover hot topics in AI in medicine as well as live questions & comments from attendees. It's casual, insightful, and open to all!

Panelists:
Akshay Chaudhari - Stanford
Hugh Harvey - Hardian Health
Sharon Zhou - Stanford
Rogier van der Sluijs - Stanford
Alaa Youssef - Stanford

Outline:
01:36 - Intros and thoughts on foundational models: balance of broad training and specific medical use, ethical, legal and consent

09:22 - Academic and Industry collaboration: UK has specific grants to support, academic cloud compute credits, limiting bias and increasing trust with transparency

15:54 - Interpretable vs usable vs explainability: who is the end user and how much do they rely on the output

23:35 - Monitoring algorithms in the wild: human computer interaction, proper reliance, proactive and reactive post market surveillance, adverse event reporting

31:07 - Characterizing robustness: defined per use case, fail compared to human metric, synthetic data to address weakness, balance of detecting race in medical imaging

45:18 - Factors to increase public trust: time, different people trust different people, mitigate bias, include ethical consideration in study design, editorial responsibility of medical journals

52:17 - Building a hospital from scratch with AI: clean/safe data collection, no paper, open internal framework, VR/AR, preventive infrastructure, expectation management/education

Stanford AIMI: https://aimi.stanford.edu
Twitter: https://twitter.com/StanfordAIMI

Видео AI Happy Hour | AIMI Symposium Q&A канала Stanford AIMI