Viewing profile — 256lie

conformal prediction

There are efforts like the UK Biobank but healthcare institutions are very sensitive about their patient data.

For breast screening, this task is high volume and low prevalence and AI can help with radiologist burnout from increased caseload.

For a breast screening application, it will always be confirmed with manual review before biopsy.

For screening, it depends on the false positives rate. A radiologist with have to check every positive prediction. Although, I believe in Europe, they have approved AI to be used a…

Even the average radiologist is high variable, not to mention inter-reader variability.

The article is simplified (a retrospective metastudy) and might not be indicative of what real-life performance. Even reader studies (which would be more rigorous) skip so much tha…

Clinical AI (which is currently regulated as a CAD medical device by the FDA) won't replace radiologists but treated as an additional clinical vendor application integrated into ex…

It's not an issue of resolution but of generalizability. Populations and scanners shift over time and the biggest issue in clinical AI is the changing data distribution, such as da…

Also publications are not what determines if AI get deployed in clinical practice. That's the job of the FDA and million of dollars spent on validation like clinical trials and qua…

It's known as automation bias and a problem in pilots as well as doctors. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7651899/

Meta analysis is more common in medical journals than computer science conferences. The evaluation of AI medical devices is determined by regulatory agencies like the FDA.

Additionally, breast cancer screening is a high-volume and low-prevalence task and CAD applications has been developed for decades (although not with the performance of latest CNN …

Would be interesting to see the time advantage. Mammography is high-volume and low-prevalence task with standards such as BI-RADS. While AI will not replace radiologists, breast ca…

This article is shallow and generic. Replace "software" with any other business activity and the strength of argument remains the same.

They mention a GAN which is a generative model and currently no good measure of evaluation (beside metrics like Fréchet inception distance). The article only mentions a qualitative…

Watson missed the DL train and IBM should have partnered with a company that had experience in getting medical devices through the FDA (like MSFT are doing with Nuance).

There are healthcare startups around fraud detection, reducing no-shows, telemedicine, drug discovery, and patient triage. Just radiology alone is prime for ML due to existing digi…

Is DNA intelligent? What about virus? Ants? Dolphins? A corporation?

How much more of a human are you than a radio? A computer?

That presumes technical people believe in AGI. I would think more ML researchers don't so just avoid the term "AI".

What would a principled reason for association look like beyond mere convention? Language is used by different groups to mean different things. Machine learning, logic, control, ro…

Is that a bad thing? Tesla's consumer base is not large or diverse enough to be representative of the national economy or population.

Humans can do a lot of other things such as hear car horns, reason about driver behavior, interpret road signs, and understand casuality of driving off a cliff.