No announcement yet.

Radiology Future Outlook

  • Filter
  • Time
  • Show
Clear All
new posts

  • Radiology Future Outlook

    I am a third year medical student pretty interested in pursuing a radiology residency. I have been following this forum for quite some time and know there are some radiologists on here. I was wondering what the future of the field looks on with the utilization of AI and how that will affect compensation and job prospects going forward.

  • #2
    I’m not Radiology but work in EM and I don’t see any way AI is going to significantly change the job market any time in the near future.


    • #3
      I work in pathology and even in my field, I don’t see AI as a threat. There are other more important issues like cuts in Medicare reimbursement that could affect compensation.


      • #4
        I think that if you want to be paranoid other fields have just as many problems ahead. Seems like few radiologists regret their choice and I doubt many would be happier in other fields. I’m interested in rads too and have 0 concern about AI.


        • #5
          There is always a threat of reimbursement cuts and increased volume. But AI is not something I would worry about. I work in a practice where we use a lot of AI platforms daily (at least 4 I can think of per shift), and it's nowhere near replacing rads. It's just another tool you can use to increase productivity and accuracy.


          • #6
            Far from replacing them, AI has increased the number of things that radiologists do.


            • #7
              AI should be viewed as a productivity tool in almost every aspect. It may cut the labor content but not the value.
              Think of it as email compared to word or type writers. The written communication still requires a human brain element.


              • #8
                - "It’s just completely obvious within five years deep learning is going to do better than radiologists.”
                - “They should stop training radiologists now.”

                These are direct quotes from a big name in the deep learning field... made in 2016. Here we are six years later. AI has a minimal role in my practice. I get frequent unsolicited calls / emails / texts from recruiters because the demand for radiologists nationwide is so high.

                I can't predict the future. But you can see how the predictions of doom have gone so far.


                • #9
                  I am retired from the practice of radiology (1996-2021). Throughout my career, there was consistent and pervasive doom and gloom. Reimbursement cuts, threats of capitation, being replaced by specialists for specialty reads, being replaced by mid-levels, being replaced by offshore rads from India, AI, loss of turf, etc. I remember one attending in residency (early 90’s) telling a bunch of us that he felt like he was training blacksmiths at the turn of the century. Despite all of the negativity, over the years the work of radiologists increased and the role became increasingly essential to hospitals and medical care in a wide variety of fields and settings The demand for radiologists in the current job market has probably never been higher than it is now. I cannot predict the future, but I believe if you think you would like radiology as a specialty, you should do it.


                  • #10
                    Fellow rad. The main difference I’ve seen over the years is productivity increase demands. Less commonly, rad take full lunch break (if they don’t have tumor board). They may be glued to the chair cranking through studies. This is to accommodate the declining reimbursements. This has actually worked in a way as incomes have steadily risen. However, everyone has a max comfort speed. After a point, the job starts to get a little exhausting. This sort of thing happens in private practice, academics, corporate, employed, telerads etc.

                    There are opportunities with acceptable volumes. But, you would need to sacrifice vacation or salary (unless in a very rural town with supplemental income from hospital or high reimbursement).

                    For better or worse, there are way more jobs that allow you to work from home even if hospital based. This took off with Covid and is not going back in my opinion. There will be a mix of in-house shifts and at home shifts. That’s appealing to some.

                    There’s also telerads, which is truly just a fraction of what a complete radiologist is. Con is you are working for a corporate group and your income per study is horrible, no way around it. Volumes have to be high.

                    Interventional Radiology is basically becoming a different specialty for many settings. The classic IR doc that can crush the stack during “free time” is trending away. There’s clinic and practice development to manage. The current training (2 IR years, 3 DR years)also doesn’t really allow IR doc to do much more than IR and maybe ED and plain fine, unless they do lots of reading at home or extra elective rotations during residency.


                    • #11
                      Issue with AI is that humans are fickle subjects in a real world setting. I would go to a program that has exposure to it, so you can know workarounds and interpretation around AI algos, though. It's not that they won't exist, it just won't replace us in the near future.

                      Clinicians are already working with their replacements (the NPs, PAs, CRNAs). The suits see them as equivalent productivity for less cost... so yea.

                      Chill job and almost guaranteed spot- Combined nucs/rads residency. Few fellowship trained nucs. There is a 5 year combined track, I believe.

                      Just remember- you can find your ideal job in any subspecialty if you have geographic flexibility. Overall, the median radiology job is still better than many others.


                      • #12
                        It will be interesting to see how far IR diverges. Right now, they learn a lot of DR that they do not need to know if they are doing full time IR. They would probably be better off with more time on IR and on clinical services and less learning how to read studies they will never read.
                        Their work is image guided but that is very different from reading studies. And many of the things DRs learn are useless for IR.

                        Our IR people are far too busy to even attempt reading any cases off the list. That is even with the DRs doing some of the lower level stuff like biopsies and drainages.
                        The people who a few years ago were claiming that Ai would put radiologists out of work had no idea what they were talking about. They were wrong about what AI can do plus they did not understand what radiologists do.

                        Many fields of medicine have discovered that much of what they do can be done remotely and that Ai can help collect and analyze information. That does not mean there will be no need for doctors.
                        If you can be replaced by a NP or PA, then most of what you are doing is NP or PA work, not practicing medicine. If much of what you do is in the overlap zone, then hire NPs or PAs to do what they can do and devote your time to medicine.

                        AI and telemedicine will change all fields, not just radiology. Predicting what they will look like in 10 years is hopeless.


                        • #13
                          If AI can read radiology studies as well as a computer can drive my Tesla, you have nothing to worry about. The computer is a great driving partner, definitely makes life better under a limited number of circumstances, but AI won't be replacing me as the driver for a long time to come.


                          • #14
                            OP, who knows?

                            My group has been talking about losing the contract since I started plus I've been chicken-little-ing about mid-levels taking our jobs for over a decade. Maybe it ends tomorrow? In the meantime, I've gone from new grad to one email away from retiring--in other words, a whole career essentially finished in the shadow of uncertainty.


                            • #15
                              Are the current AI devices in the radiology world any good or providing any value?