What’s the worst thing about radiology?

This forum made possible through the generous support of SDN members, donors, and sponsors. Thank you.

Steve_Zissou

Full Member
7+ Year Member
Joined
Aug 4, 2014
Messages
5,951
Reaction score
11,118
Howdy everyone. I’m a second year Med student that is currently scoping out potential specialties. I was looking at anesthesia for a long time, but the current situation regarding CRNAs and the battle for independence is turning me off to the specialty. A mentor recommended I look into radiology and so I wanted to ask you all, what is the worst (or few worst) thing about radiology? What’s the biggest worry in the field and for the future right now?

Members don't see this ad.
 
I would say the worst part of Radiology is discussing the threat of AI with med students and undergrads who are convinced it will replace us in the next few years.
 
  • Like
  • Haha
Reactions: 25 users
Members don't see this ad :)
I would say the worst part of Radiology is discussing the threat of AI with med students and undergrads who are convinced it will replace us in the next few years.
Anecdotally, I feel like the majority of people who have this idea are not in/involved in radiology.
 
  • Like
Reactions: 1 user
Classic Dunning-Kruger effect - those who understand radiology least (aka non-radiologists) feel the most qualified to discuss AI taking over radiology. It is nothing new, along the same lines of non-radiology physicians thinking they know how to read CTs well - you don't know what you don't know. What a boring topic. Midlevel encroachment on all the other specialties is a much more real threat than whatever theoretical fantasies people seem to have about AI capabilities in radiology.

But to get back to the topic, my least favorite thing about radiology is how many radiologists are happy to work like dogs to earn higher salaries. I would rather have a nice lifestyle and earn less money and enjoy my job, but the breakneck pace of modern radiology makes it less enjoyable. Many radiologists seem to actually prefer that, but it is a big turn-off to me about the field. I will be in this field for several decades; enjoying my work-life through those decades without feeling like I am drinking from a firehose at work every day seems far more important to me than making as much money as I can.
 
  • Like
Reactions: 8 users
Classic Dunning-Kruger effect - those who understand radiology least (aka non-radiologists) feel the most qualified to discuss AI taking over radiology. It is nothing new, along the same lines of non-radiology physicians thinking they know how to read CTs well - you don't know what you don't know. What a boring topic. Midlevel encroachment on all the other specialties is a much more real threat than whatever theoretical fantasies people seem to have about AI capabilities in radiology.

But to get back to the topic, my least favorite thing about radiology is how many radiologists are happy to work like dogs to earn higher salaries. I would rather have a nice lifestyle and earn less money and enjoy my job, but the breakneck pace of modern radiology makes it less enjoyable. Many radiologists seem to actually prefer that, but it is a big turn-off to me about the field. I will be in this field for several decades; enjoying my work-life through those decades without feeling like I am drinking from a firehose at work every day seems far more important to me than making as much money as I can.
Would you mind sharing your typical hours and rough ballpark average amount of reads you do in that given time as an attending?

This number seems to be all over the place.
 
Last edited:
  • Like
Reactions: 1 users
Don't choose your specialty based on perceived threats to the profession chicken little.

Definitely not, but I do want to compare the down sides of each of the specialties I’m interested in. I’ve heard the pros from mentors who love their field, now I want to hear the real cons.

Regarding AI, I have no worries about that. After reading Consciousness Explained in which the author goes fairly in depth into the development of AI, I have to imagine it’ll be a long time before AI replaces high-skill workers.
 
Would you mind sharing your typical hours and rough ballpark average amount of reads you do in that given time as an attending?

This number seems to be all over the place.

The number is going to be all over the place because it has so many variables. TBH it could be anywhere from 25-150.
 
  • Like
Reactions: 1 users
i will offer a contrasting opinion to the above: only pursue radiology if you cannot see yourself happy in any other field

the world is changing fast. and you are a second year medical student. you are nearly a decade removed from becoming a practicing radiologist.

i think radiologists are inherently defensive when it comes to the threat of AI to their field. the reality is that it is a very real threat to disrupt the field. other fields are not immune to AI, but those same fields are definitely more resistant

for example, i think COVID-19 is running a nice simulation of what is to come in regards to how decreased volume will disrupt the rads workforce. let me elaborate: a good portion - maybe most? - of radiology studies ordered are NORMAL. when (not if) the big players in AI (google, microsoft, facebook, etc backed by unlimited $$) automate NORMAL reads, the amount of imaging that needs to be reviewed by human eyes drops significantly and the efficiency of a single radiologist increases exponentially. this leaves less and less work to be done by human radiologists, and thus will lead to a massive oversupply of rads

what happens to the rest of rads workforce? do they re-train? re-enter a new field altogether? idk. but i do know that most of the younger rads will get screwed as they will have graduated with immense loans but no big paychecks to pay them off. imo, i'm not sure its worth the risk.
 
  • Like
Reactions: 1 users
i will offer a contrasting opinion to the above: only pursue radiology if you cannot see yourself happy in any other field

the world is changing fast. and you are a second year medical student. you are nearly a decade removed from becoming a practicing radiologist.

i think radiologists are inherently defensive when it comes to the threat of AI to their field. the reality is that it is a very real threat to disrupt the field. other fields are not immune to AI, but those same fields are definitely more resistant

for example, i think COVID-19 is running a nice simulation of what is to come in regards to how decreased volume will disrupt the rads workforce. let me elaborate: a good portion - maybe most? - of radiology studies ordered are NORMAL. when (not if) the big players in AI (google, microsoft, facebook, etc backed by unlimited $$) automate NORMAL reads, the amount of imaging that needs to be reviewed by human eyes drops significantly and the efficiency of a single radiologist increases exponentially. this leaves less and less work to be done by human radiologists, and thus will lead to a massive oversupply of rads

what happens to the rest of rads workforce? do they re-train? re-enter a new field altogether? idk. but i do know that most of the younger rads will get screwed as they will have graduated with immense loans but no big paychecks to pay them off. imo, i'm not sure its worth the risk.

My understanding was that it is very unlikely that AI will do that within my lifetime (mid-20s here).

Honestly, radiology so far has shown the biggest pros with the fewest cons for me so far. I really like the idea of being an expert in whatever I do, being able to diagnose and move on, and working with technology. The overall lifestyle doesn’t seem bad for having a family and raising kids either.
 
  • Like
Reactions: 1 users
i will offer a contrasting opinion to the above: only pursue radiology if you cannot see yourself happy in any other field

the world is changing fast. and you are a second year medical student. you are nearly a decade removed from becoming a practicing radiologist.

i think radiologists are inherently defensive when it comes to the threat of AI to their field. the reality is that it is a very real threat to disrupt the field. other fields are not immune to AI, but those same fields are definitely more resistant

for example, i think COVID-19 is running a nice simulation of what is to come in regards to how decreased volume will disrupt the rads workforce. let me elaborate: a good portion - maybe most? - of radiology studies ordered are NORMAL. when (not if) the big players in AI (google, microsoft, facebook, etc backed by unlimited $$) automate NORMAL reads, the amount of imaging that needs to be reviewed by human eyes drops significantly and the efficiency of a single radiologist increases exponentially. this leaves less and less work to be done by human radiologists, and thus will lead to a massive oversupply of rads

what happens to the rest of rads workforce? do they re-train? re-enter a new field altogether? idk. but i do know that most of the younger rads will get screwed as they will have graduated with immense loans but no big paychecks to pay them off. imo, i'm not sure its worth the risk.

Oh look, another non-radiologist here to offer their opinion. :rolleyes:
 
  • Like
Reactions: 4 users
Pros: Pretty decent $$$, less paperwork BS compared to other specialties, minimal patient interaction, clinicians really rely on what you say, anatomy is interesting

Cons: Work is not chill, call is insane, significant nights and weekends, you are a cog in a machine
 
  • Like
Reactions: 3 users
Pros: Pretty decent $$$, less paperwork BS compared to other specialties, minimal patient interaction, clinicians really rely on what you say, anatomy is interesting

Cons: Work is not chill, call is insane, significant nights and weekends, you are a cog in a machine

Who isn't at this point?
 
  • Like
Reactions: 1 users
Members don't see this ad :)

I'd be interested in Urology or ENT since the ones I've met seem to have a decent work/life balance. All the general surgeons, ortho, and plastic surgeons I've met say it's their life and they have a difficult time keeping up with anything outside of surgery. A ortho my mom works with quite directly told me "don't do ortho, you won't have time for the people you love."
 
  • Like
Reactions: 1 user
Pros: Pretty decent $$$, less paperwork BS compared to other specialties, minimal patient interaction, clinicians really rely on what you say, anatomy is interesting

Cons: Work is not chill, call is insane, significant nights and weekends, you are a cog in a machine
My issue is that if you showed me roughly that list and said it was for anesthesia I would agree with it. I'm having a lot of trouble choosing between the two even after a radiology rotation.

If OP doesn't mind, I would ask a follow-up question. Do good private practice radiologists come out of below average (read not academic) residency programs if they put in the work? If one's goal is to be a great PP radiologist, is one noticeably hampered by the quality of their training than other fields?
 
  • Like
Reactions: 1 user
Great radiologists come from great and ****ty programs. ****ty radiologists come from great and ****ty programs.

If you want PP you do not need a top program. In some ways it's better to go to a workhorse mid tier program.
 
  • Like
Reactions: 2 users
Yikes, if a day in the life of a radiologist isn't relatively chill, what is? Even on my derm and ophtho blocks, clinic for them was decidedly not chill either, with ~40 patients scheduled in 8 hours. None of the surgeons seem to be chilling either :/

What's a lifestyle oriented MS3 to do?
 
  • Like
Reactions: 1 user
Yikes, if a day in the life of a radiologist isn't relatively chill, what is? Even on my derm and ophtho blocks, clinic for them was decidedly not chill either, with ~40 patients scheduled in 8 hours. None of the surgeons seem to be chilling either :/

What's a lifestyle oriented MS3 to do?
Allergy fellowship
 
  • Like
Reactions: 6 users
Yikes, if a day in the life of a radiologist isn't relatively chill, what is? Even on my derm and ophtho blocks, clinic for them was decidedly not chill either, with ~40 patients scheduled in 8 hours. None of the surgeons seem to be chilling either :/

What's a lifestyle oriented MS3 to do?
It depends on how much money you want to make. There are rads jobs that are relatively chill. Find of private practice group that values lifestyle over maximizing income, work at the VA, do telerads, or work in academics in a clinical position without getting caught up in the publishing rat race. You'll obviously take a pay cut for all those options but plenty of people thing it's worth the trade off.
 
  • Like
Reactions: 3 users
SOME surgeons can avoid being cogs in the machine by embracing the business aspect of things and working harder than most of their peers. The problem is many (most?) new grads are allergic to anything that requires risk and totally embrace the mindset of clock-in, clock-out. I think it's important to distinguish that going into ENT/Uro/Ortho etc may make it easier to have some autonomy, but simply completing residency does not make it so. Whatever you choose you have to work very hard to make your own path independent from the suits. Rads can make it work as well, though it may be harder.

It's also been 4 years since Geoffrey Hinton, godfather of machine learning, said to stop training radiologists because they would be obsolete in 5 years. Well only 8 months to go until radiologists are replaced :rolleyes:
 
  • Like
Reactions: 7 users
Allergy fellowship
Thats my desired specialty. Every Allergist Ive shadowed or have a relationship with LOVES their job. The path is interesting (at least to me) and they said their lifestyle is amazing. Literally work a school day (8-3ish) 4 days a week and one day a week go in half day for paperwork. Most if not all of them make 325K+ that have mentored me
 
i will offer a contrasting opinion to the above: only pursue radiology if you cannot see yourself happy in any other field

the world is changing fast. and you are a second year medical student. you are nearly a decade removed from becoming a practicing radiologist.

i think radiologists are inherently defensive when it comes to the threat of AI to their field. the reality is that it is a very real threat to disrupt the field. other fields are not immune to AI, but those same fields are definitely more resistant

for example, i think COVID-19 is running a nice simulation of what is to come in regards to how decreased volume will disrupt the rads workforce. let me elaborate: a good portion - maybe most? - of radiology studies ordered are NORMAL. when (not if) the big players in AI (google, microsoft, facebook, etc backed by unlimited $$) automate NORMAL reads, the amount of imaging that needs to be reviewed by human eyes drops significantly and the efficiency of a single radiologist increases exponentially. this leaves less and less work to be done by human radiologists, and thus will lead to a massive oversupply of rads

what happens to the rest of rads workforce? do they re-train? re-enter a new field altogether? idk. but i do know that most of the younger rads will get screwed as they will have graduated with immense loans but no big paychecks to pay them off. imo, i'm not sure its worth the risk.

The problem is there are some pathologies and anatomic variants that are rare enough that there simply isn’t a large enough dataset to teach a CNN to identify it reliably. If you’re teaching the CNN to call any sufficiently different-from-“normal” image with high sensitivity, there are going to be a LOT of normals that are called abnormal.

Put another way: you can teach a monkey to do surgery. The surgeon is REALLY just insurance in case something goes wrong enough that you need improvisation, which only comes with enough high-end experience.

You can teach a monkey to read images, but when the image doesn’t fall into a neatly pre-defines category, that’s when you need a radiologist to give their input. And, simply put, there are so many diseases which present on imaging in such rare ways that you simply don’t have enough data to teach the AI “this isn’t normal.”

The reason AI won’t replace radiologists to a large extent comes down to the same reason Quicken wont replace accountants. You’re severely overestimating the extent to which things in the field can be neatly categorized, even if those categories are just “normal” and “abnormal.”
 
Last edited:
  • Like
Reactions: 6 users
It depends on how much money you want to make. There are rads jobs that are relatively chill. Find of private practice group that values lifestyle over maximizing income, work at the VA, do telerads, or work in academics in a clinical position without getting caught up in the publishing rat race. You'll obviously take a pay cut for all those options but plenty of people thing it's worth the trade off.
How much of a pay cut, I guess is the question. I don't mind living in the Midwest in some flyover state (I did college in Missouri) which should help. From reading these boards, it looks like there are plenty of private practice groups out there that "value lifestyle" and get >500k as partner doing M-F shifts (plus occasional evening or weekend call shift).

I can live with that, hopefully that will still be the case in 6-7 years when I'm on the market.
 
  • Like
Reactions: 3 users
Thats my desired specialty. Every Allergist Ive shadowed or have a relationship with LOVES their job. The path is interesting (at least to me) and they said their lifestyle is amazing. Literally work a school day (8-3ish) 4 days a week and one day a week go in half day for paperwork. Most if not all of them make 325K+ that have mentored me
Friend's father is an allergy doc in private practice. Said his clinic does a lot of those desensitization shots and has been making >500k with fantastic lifestyle. Seems like the MGMA averages aren't that accurate for business minded docs with some hustle.
 
  • Like
Reactions: 2 users
Friend's father is an allergy doc in private practice. Said his clinic does a lot of those desensitization shots and has been making >500k with fantastic lifestyle. Seems like the MGMA averages aren't that accurate for business minded docs with some hustle.
Yeah you can do super well in Allergy especially with a business mind. Do you need a subscription to MGMA to get the data? I can't seem to find it. I am curious to see what they say for Allergy
 
Yeah you can do super well in Allergy especially with a business mind. Do you need a subscription to MGMA to get the data? I can't seem to find it. I am curious to see what they say for Allergy
Yeah you normally have to pay several hundred dollars for the MGMA access

PM me your email and I can send you that
 
  • Like
Reactions: 1 user
The worst thing about radiology is that radiologists only make 450-600k/year :p.
 
  • Like
Reactions: 1 user
If you’re single and stay that way, you can have a very nice lifestyle with a very chill radiology job.

I doubt there are many debt-free single rads out there hating their jobs because they can’t say no to grinding it out more and more.
 
If you’re single and stay that way, you can have a very nice lifestyle with a very chill radiology job.

I doubt there are many debt-free single rads out there hating their jobs because they can’t say no to grinding it out more and more.
Doesn't partnership anchor people? Dropping my salary hundreds of thousands for a few years seems like a significant hurdle to changing groups?
 
People are worrying about AI when the worst thing about radiology is the cyclic nature of its job market. People are already talking about hours being cut and jobs delayed and how it's gonna a be a tough market for new grads in the next couple years.
 
Have to say, I'm on a remote radiology elective right now and they've given us a little reading about AI. Some stuff out there looks pretty impressive, if not "replacing" then certainly disrupting. One paper from a few months ago showed a neural net can correctly identify >70% of normal CXRs as normal with high specificity.

Obviously a radiologist will be needed to describe exactly how an abnormal xray is abnormal for a long time, but normal imaging is a significant part of daily volume, isn't it? If we can start trusting AI to filter out normal imaging as normal anytime soon, that's a huge drop in demand.
 
If the specificity is anything less than 100% it won’t help at all. Calling a study normal is the most dangerous thing a radiologist can do because he shoulders all of the potential malpractice responsibility.

I should rephrase: if the specificity is any less than it would be with a combined radiologist:AI analysis, it won’t increase the volume rate at all.
 
Last edited:
  • Like
Reactions: 1 user
Well, really it just has to be non-inferior to the average radiologist, no?

But just in terms of workflow and prioritization, if you had a tool that could flag Normal with >95% specificity, that wouldn't be appealing to a practice?
 
I have been impressed by some of the Head CT bleed detection software with 95 plus sensitivity and specificity. I think it will be a useful tool in detecting findings and the interpreting physician will have an added security that he/she did not miss the finding and may even make them more efficient. Technology will continue to improve . Currently we are on the apple II e version. We have a ways to go, but it will help all of medicine not just radiology.

In the Covid era, it would be amazing if we could use it to optimize vent settings and predict who is going to do well and who is not and how to interpret the various labs LDH, ferritin, ESR/CRP/ d dimer/ fibrinogen etc and the ABG and stratify these patients prognosis and identify optimal treatments.
 
Well, really it just has to be non-inferior to the average radiologist, no?

But just in terms of workflow and prioritization, if you had a tool that could flag Normal with >95% specificity, that wouldn't be appealing to a practice?

I’m a patient that developed metastatic bronchogenic carcinoma that, in hindsight, was surgically resectable as seen on a routine outpatient CXR except that the AI didn’t notice. However, it was in hindsight visible to the human radiologist—it just so happened that nobody bothered to take a look at it because the AI flagged it normal, and the AI had 99% specificity so that was thought acceptable.

I would definitely be justified in suing because of this. Only takes a few such cases for practices to double check all the AI ”normals” with human eyes, even if 99% of the time it isn’t necessary. As per irwarrior, in that case the combined AI:radiologist read would probably have greater sensitivity and specificity than the radiologist alone... but it would also have greater specificity and sensitivity than the AI alone, in which case the standard of care becomes a combined AI:radiologist read, increasing accuracy, but not saving that much time.

Put another way, for AI alone to become standard of care, it has to be noninferior to the AI:radiologist combo, not just the average radiologist.
 
  • Like
Reactions: 2 users
Perhaps that is the current best algorithm AI+ Radiologist>Radiologist alone>> AI alone.

My worry is that we rely on AI as a crutch and stop learning the radiology like we used to. It would be similar to the calculator being now a necessity and humans are not as good at math without it. This is why we should train without AI first and then use it only after getting a solid foundation. It would be tricky if we make AI the new gold standard (better than your expert faculty) when it yet has to be proven. Not saying things won't change, but right now it is a fascinating adjunctive tool as opposed to a replacement in most places.
 
Perhaps that is the current best algorithm AI+ Radiologist>Radiologist alone>> AI alone.

My worry is that we rely on AI as a crutch and stop learning the radiology like we used to. It would be similar to the calculator being now a necessity and humans are not as good at math without it. This is why we should train without AI first and then use it only after getting a solid foundation. It would be tricky if we make AI the new gold standard (better than your expert faculty) when it yet has to be proven. Not saying things won't change, but right now it is a fascinating adjunctive tool as opposed to a replacement in most places.

Step 1. Fire 3/4 of radiologists
Step 2. Have the ones left oversee 4 mid levels with AI
Step 3. Profit???
 
Step 1. Fire 3/4 of radiologists
Step 2. Have the ones left oversee 4 mid levels with AI
Step 3. Profit???
Interesting. What’s the difference between overseeing AI and doing the work of a typical radiologist. I am sure anesthesia thought there would be a less demand for them with CRNAs but there is still an increasing demand bc they can’t a) do better work than the actual anesthesiologist and b) the case for surgery will continue to rise. I see similar parallels with this to AI radiology. With surgeries just being replaced by imaging. Could be wrong but just a thought. ¯\_(ツ)_/¯
 
Anesthesia has increased its space in the non OR environment (Cath lab(structural hearts)), VIR and neuroIR, Endoscopic suites. CRNAs are able to practice independently in some states/hospitals.

AI will be a game changer for a lot of industries in and out of medicine and radiology, but it is not quite ready for prime time outside of select indications.
 
  • Like
Reactions: 1 user
Howdy everyone. I’m a second year Med student that is currently scoping out potential specialties. I was looking at anesthesia for a long time, but the current situation regarding CRNAs and the battle for independence is turning me off to the specialty. A mentor recommended I look into radiology and so I wanted to ask you all, what is the worst (or few worst) thing about radiology? What’s the biggest worry in the field and for the future right now?

Getting drafted back on the medicine floor during an pandemic.
 
  • Like
  • Haha
Reactions: 2 users
Getting drafted back on the medicine floor during an pandemic.
I went back to the medicine floor and it was not that bad. The ennui of staying at home is worse.

The pandemic allows us to remind everyone else in medicine that radiologists are also doctors and we are standing with them.

A mentor recommended I look into radiology and so I wanted to ask you all, what is the worst (or few worst) thing about radiology?
When other doctors don't treat you as a fellow doctor and consultant, such as by ordering studies over your objection that they don't make any sense, or ignoring your assessment without a discussion with you.
 
Last edited:
  • Like
Reactions: 2 users
Anesthesia has increased its space in the non OR environment (Cath lab(structural hearts)), VIR and neuroIR, Endoscopic suites. CRNAs are able to practice independently in some states/hospitals.

AI will be a game changer for a lot of industries in and out of medicine and radiology, but it is not quite ready for prime time outside of select indications.
Even with CRNAs getting to practice independently, there is still a demand for anesthesiologists. I am sure the good radiologist will adapt to their new environment. Plus I would judge someone on their behavior. If radiologists thought they would be out of work, they would pick a new field yesterday.
 
  • Like
Reactions: 1 user
Anesthesia has increased its space in the non OR environment (Cath lab(structural hearts)), VIR and neuroIR, Endoscopic suites. CRNAs are able to practice independently in some states/hospitals.

AI will be a game changer for a lot of industries in and out of medicine and radiology, but it is not quite ready for prime time outside of select indications.

...yet

the fact is that current and future trainees need at least 20-30s to re-coup their med school investment. that's a long, long time for AI to progress. you can't deny that there are fields (clinical and procedural) much safer than rads.

my advice to med students - proceed with caution
 
Top