I’m just watching lecture 14 where @jeremy talks about Children’s Hospital LA: “When the doctors do rounds, data scientists come with them. I don’t know anywhere else in the world where this happens.”
Funny coincidence, earlier this year I discussed about somewhat similar project in Finland. The children’s hospital in Helsinki and IBM are working to utilize Watson in monitoring small premature infants. The project member told me that an important part of their job was to actually go with the doctors and try to learn what to model for Watson. She was pretty excited about the project and told me they have found ways to improve the process at the hospital, even though the hospital had a very good process to begin with. It looks like there’s room for improvement even in well known use cases.
The only reference I found is in the hospital’s annual report: “Nurses and physicians recognize symptoms through experience, of course. The point of Watson is to pick up on connections that humans cannot necessarily detect”, says a chief physician.