>Diagnosis is classification.

A radiologist's diagnosis is not image classification, it's reality classification maybe. (That's quite poetic).

Watching an educational Youtube video about endangered tiger habitats is not the same thing as segmenting possible embedded pictures of kitties and poachers or whatever, and classifying them as such. There's, like, a lot of additional context.

To flesh this out a little more:

A (good) radiologist is interpreting the images in light of the patient's history, symptoms, and other tests, with the goal of forming a diagnosis and treatment plan. This is rather different from taking an MxN array of pixels and trying to decide if it contains a tumor.

For example, about 10% of people have small gallstones. If an ultrasound incidentally detects some in a healthy, asymptomatic person, nothing happens. The exact same images, coming from a patient with a history of upper-right abdominal pain and jaundice, probably lead to a referral for surgery instead.

What you described is outside the scope of the problem being solved in radiological classification. That problem- like protein structure predction- is an intentionally simplified process used to make it possible to fairly compare humans vs ML.

What you're describing is the general problem of informed diagnosis, which is also classification, but typically takes into account a great deal of qualitative information. Few if any ML people are working in this area because there is no golden data and it's nearly impossible to evaluate in a quantitative way.

Agreed--but that's the point: the demand is not actually for radiological classification, it's for accurate, actionable diagnoses. Heck, radiologists call what they do "interpretation."

I can certainly imagine that radiologists would love (say) a tool that automatically flags low-quality or mis-oriented images, but that's not at all where the hype is at.

Those things already exist. It was very enlightening- my uncle was a radiologist and I spent a few hours watching him do his job. The software they use is extremely sophisticated with lots of custom bells and whistles (and the monitors have insane contrast ratios). Most people don't see it but non-ML medical imaging is extremely mature and is developed in close contact with the users.