why is radiology getting more competitive even with the uncertainty from AI

[PubPhD]

The real goal of AI is not to create cheaper health care for '"he sake of the patient"

Just as you and the Money Moniker have differing views on what it is to practice medicine, I assure you that while some people do want to save administration money (IBM Watson), the vast majority of us ML researchers are doing it purely to help improve the lives of others. The ML PhD salaries are sweet, but I'd pull the same hours doing the exact same thing for 60k/year as I feel this is the most effective way I can help the most people.

That being said, I'm all for saving money by reducing patient demand if the case can be handled more efficiently by a model and doesn't require a physician. However, this would be in systems like the UK or Canada where the growing population is placing a burden on the national healthcare system.

---

Members don't see this ad.

 

[Money Moniker]

The goal of medicine is not to profit off sick people. Radiology is a part of medicine.
In practice, we end up doing a lot of free work for patients, whether we like it or not. If you're not comfortable with that, you may want to choose another profession.

What I object to is a shifting of the money flow. The real goal of AI is not to create cheaper health care for '"he sake of the patient", but divert the money to corporate and tech stakeholders. You'll be stocking shelves for the sake of the non-profit's profit margin.

If you didn't get a paycheck, would you show up to work "for the sake of the patient" or would you instead find another line of work that provided you with money? I'm guessing you'd stop showing up to work and helping patients if you were not getting paid for it, which means the reason you treat patients is to make money and not out of altruism. This describes basically 99.9999% of humanity across the spectrum of economic activity and is nothing to be a ashamed of nor dissimulate about.

Technically, it would be great for the patient if I worked for free and lived in his dog-house for easy accessibility, but I am opposed to such an outcome, even if my opposition is "bad" for the patient given the "great" alternative. Reason I oppose it? Because it's bad for me. Likewise, it may be great for the patient if a robot could do my job and I was instead forced to work at Walmart for $8/hour, but again, I am opposed to this outcome because while good for the patient, it's bad for me. I could do this all day but I think you get the point. We're doctors, but this doesn't mean we are supposed to give up our humanity and subordinate every last ounce of self interest under "what's good for the patient" else be become literally subhuman slaves.

 

[Cognovi]

Honest question, whose responsibility should this fall under? In my field, an integration of recent research is outside of our domain as it does not contribute anything new. My opinion is that since these techniques can significantly improve patient care/reduce ER load, wouldn't most medical professionals push to integrate this as fast as possible for the sake of the patient?

You should find a radiologist to collaborate with to do a pilot of clinical integration, who can write up that experience for a clinical radiology journal.

 

[Puff-of-Snow Sign]

Mid-levels. Much cheaper for an established derm practice to hire one rather than a new dermatologist...last time I went to a dermatologist I was being seen by their NP until they found out I was a MD. Also was asked by the receptionist if it was ok if the "medical" student (NP student) was allowed to observe

Steroid cream vs. biopsy. Can't decide? Do both. Derm solved.

 

[Shufflin]

The day AI can negotiate its own salary is the day we should all cower in great fear.

 

[Taurus]

AI, radiology and the future of work

I wouldn't worry about AI in radiology. I remember a few years ago when medical students were running away from radiology because of the job market. I said at the time, buy low, sell high. Great opportunity to get into a better residency than you could otherwise.

 

[Cognovi]

AI will first automate the cognitively simple and tedious tasks. I would look forward to that. Fields without it will be left behind doing monkey tricks.

 

[808s&heartbreak]

AI will first automate the cognitively simple and tedious tasks. I would look forward to that. Fields without it will be left behind doing monkey tricks.

Hope so, but don't think it can even do that. It takes only a few hours to train a neural network to diagnose pneumothorax on hundreds of thousands of CXRs. Even that it cannot do reliably. Enlitic has been working on it for 3 years....still nothing. Radiology is way harder than these AI "experts" thought it would be.



Don't drink the kool-aid.

 

[Radology]

That video is amazing. The reporter asks a radiation oncologist if AI reads scans better than him, seeming to not even realize that radiology is a field.

 

[Cognovi]

The video's last claim is that their algorithm increases sensitivity of detection of cancer. No mention of specificity.

 

[Radology]

Can’t miss a cancer if you say that everyone has cancer