I disagree with all your points. I think the main reason for differences in salary is that it is easy to measure what a radiologist does. It is also relatively

I disagree with all your points. I think the main reason for differences in salary is that it is easy to measure what a radiologist does. It is also relatively easy to measure what a cardiologist does, particularly if they are doing procedures such as angiograms. Mostly a neurologist diagnoses stuff and manages chronic disease, and how do you measure that? Arguably, the neurologist does the most difficult and complicated job. They get paid based on consultations per time, which is clearly a pretty hopeless measure. Also, the notion that radiology is going to be automated by increased processor power is a bit ridiculous. We are talking about a task where lives are at stake, communication and DISCUSSION with real people is required frequently, and there are large variations in the quality and modality of scanning. Also, any data driven approach is very limited... what happens when a new MRI sequence is developed? We have to wait 15 years for real radiologists to do enough reporting so we can get a proper dataset?