61–70 of 134 posts
Earlier quoted context omitted.
This is not how trust works
You mean it's not enough to trust in average accuracy, you want trust in specific instances of application?
Earlier quoted context omitted.
They could be actually better and much faster than humans for initial diagnosis. This has far greater implications, e.g. early detection actually reduces big burden from system.
That has been available since the 1970s. Healthcare is not a very complex compsci issue for 80% of cases. The problem is presentation of symptoms is highly subjective and needs highly complex interpretation that no AI will ever be able to achieve and deal with the liability conundrum
The cost blowout in healthcare isn't due to diagnosis or care but administrative costs. All these diagnostic AIs won't significantly reduce the cost of healthcare nor improve quality. What we really need is automated administration and billings to really move the needle and that's more human bottlenecked than technology.
The cost blowout is a US thing, diagnostic AIs can be used anywhere. And even if the prices weren't affected (which doesn't seem plausible), this can hugely increase the availability and the timeliness of a reliable diagnosis.
https://outline.com/EpGqWm
Earlier quoted context omitted.
you should still complain, this nonsense has to end.
I’ll bet you say the same thing when you are forced to pay for groceries before eating them.
It's the most disingenuous argument against free media (or piracy etc) that's being thrown around.
AI is being developed to make MRI scans 10x faster. https://www.forbes.com/sites/samshead/2018/08/20/facebook-ai... If DeepMind is trained on millions of MRI’s, we might have better preventative medicine.
I have talked at length with several people in the FastMRI project and in my opinion it is actually very dangerous. There is no feasible way to validate that such a model will not hallucinate normal tissue in the presence of a rare abnormality. The argument often used is that advanced reconstruction techniques such as compressed sensing have not required validating against very rare abnormalities, however when deep n…
I view the whole issue stochastically, with the immediate aim being to make (significantly) fewer errors than the current approach, which is having a human decide. I don't claim that I can design an experiment which could selve as an indication whether we are improving upon human judgments, but I think this should be the goal.
Reflecting upon my view, i think it comes from the experience of training ml-algorithms. You are always minimizing errors, but you goal is almost never to make 0 errors, because often your data is noisy and you are probably overfitting. I know the medical enviroments are more sensitive, but I can't really wrap my head around how we could design a learning algorithms that does not make any error and works on all abnormalies. I think it will always missclassify.
Rephrasing my argument: I think the approval should be given if an significant expected improvement over the distribution of real-life abnormalies can be detected and not over the uniform-dsitribution over all abnormalies.
EDIT: detecting out-of-distribution samples is hard and I don't think this is a solution and leads to a false sense of security.