Date: December 5, 2016
- Wrong results, misinterpreted results, unreproducible results, no results.
- And if doctors don’t do it, learn it, get good at it, value it … is it useful to say that machines are better at it?
- Rule 1 – is it a task human doctors do, done with the same inputs.
- While this isn’t necessarily “outperforming” human doctors, it probably costs under a cent per patient to run the model.
- Just because something sounds medical, and seems to be in the scope of medical practice, it doesn’t mean doctors are actually doing it.
Related NLP Articles
Build smarter apps with our natural language processing API.