[DanielQuilli]

I could read zero screeners in a day and my day would still really busy with work-ups and biopsies. I feel like I need AI to make my job sustainable, hardly worried about it replacing me.

[Unknown Member]

Absolutely. Being on-site doing diagnostic workups, doing biopsies, wire locs, is a HUGE plus for job security. How many radiologists within driving distance in a given city can they call on to do what you do in person as a breast imager?  Depends on size of the city, but maybe 5, 20, 50?   As an MSK reader reading from work or home, how many radiologists could potentially provide that interpretation?  Any radiologist with a state license, which is to say, any group or individual offering remote reads so long as their docs did the paperwork to get the state license. I don’t know a number but it is a big number. 
 
IR, breast, and dedicated nucs rads (therapies, RSO, go-to MD for nuc med troubleshooting) will be relatively safe from AI in my opinion. 

[spc44_369]

Is nuclear medicine its own fellowship or a subset of body?

[Unknown Member]

You wanna sell yourself as a real nucs guy, Id say get a nucs fellowship and be good at it. This is referring to those doing DR followed by nucs fellowship , not this nuclear medicine residency stuff.

[BagsOfMilk]

Awesome, thanks for replies. Any additional thoughts on demand and salary of breast imaging in private practice now and in the future? What about combined breast 1 year fellowships?

[tdetlie_105]

Quote from BagsOfMilk

Awesome, thanks for replies. Any additional thoughts on demand and salary of breast imaging in private practice now and in the future? What about combined breast 1 year fellowships?

 
Combined fellowship is a smart move.  Unless you want to pursue high-end academics, 6 month breast fellowship would suffice.
 
Salary for breast imaging in PP will be dependent on the set-up of the group.  Partner breast imager that also takes call, does general versus an exclusive M-F non-partner breast imager will have different deals.
 
Hard to predict future demand. This is pure speculation however I can see a scenario where AI+mid-level “pre-read” a boat-load of screeners that are signed off in bulk by the breast imager or some other random radiologist if need be.
 
Procedures can be done by mid-levels which would leave diagnostics to breast imagers. If both of these scenarios actually occur, it’s logical that the demand for breast imagers would decrease.  

[Unknown Member]

If you are the primary breadwinner, which is the politically correct way of saying the man, I recommend that a breast radiologist do around 50% general at the start of his career so that he can take call like everybody else. This will ensure that you have the option of becoming (earning as) a partner if you wish.

If you are not the primary breadwinner, and you have other life responsibilities, which is the politically correct way of saying many women, it may make sense to take a M-F no evenings no weekend breast employed position that will be mostly or all breast. The extra effort to take evening and overnight call and weekends is not worth it financially for many in that situation.

[consuldreugenio]

Breast imagers are likely secure for 100% breast imaging jobs for at least the next 10 years. Either employed or PP. Breast fellowship trained surgeons and breast oncologists have made breast into a different world. It can be a different kind of pain. Someone to dump complex consults, MRIs and tumor boards onto will be hired quite quickly. General rads are washing their hands off from that these days.

I do not find 100% breast anywhere close to an enjoyable set up as a radiologist though. Mastering other subspecialties is much more fun and guarantees lifetime job security IMO.

[toumeray]

Demand for breast is high and likely always will be, even with a high performance AI. Theoretically a high confidence negative (HCN) AI product would be most useful for non breast rads reading screeners, though most likely just for reducing misses and maybe mildly increased productivity. Still need more or less exactly the same on-site breast rads as we do now, for diags, biopsies, locs, etc. If we are talking about remote screener readers, yes that may be affected but thats a totally different model and realistically often is a totally different job than the onsite breast specialist models.

As for combined, if you are for sure going private I see it as a definite plus. Im general private practices dont care whether you did a dedicated fellowship or split fellowship – you read whatever youre comfortable reading and occasionally some stuff you arent comfortable with you try to learn.

But, Id youre looking at academic centers, yes they may snub their ivory tower noses at a split fellowship person for their subspecialty academic department. But demand is so high for breast and even non breast (though especially for breast) that they would probably still take you, just talk behind your back about how youre not a real breast rad.

[6541165]

How would you all feel about someone with a 4th residency year “mini fellowship” in breast — would that count for anything? 
 
If I do mini-fellowship in breast and fellowship in combined thoracic/body I’d feel like a decent jack of all trades. Although maybe that’s not what you look for in a partner. 

[Unknown Member]

Depends on the group. Many groups are happy with having people who didnt do breast fellowship but comfortable and enthusiastic about breast imaging do it. Some groups want someone who has a breast fellowship. Most non breast fellowship people think of breast the way they do chest : as long as you are doing it and no one is complaining, its all good. As a fellowship trained breast imager, I see a range of how good radiologists are at breast – a fellowship helps, and some who have been doing it for a decade or more suck at it and no one cares except maybe the breast surgeons. Plenty of radiologists who are great at breast and do not do a breast fellowship. A mini fellowship may or may not be a real fellowship depending on institution.

[6541165]

Thanks for the feedback. Breast resident education sucks at my program so safe to assume a breast mini-fellowship would as well. 

[Unknown Member]

You can think you are a good Jack of all trades without any fellowship at all. Whether you get hired is the question, and in this market youll likely get hired. Whether others you consider you a good Jack of all trades is another question. Some dont care. Others want their performance to be respected by their peers.

[Unknown Member]

1.  Do a fellowship in something you like and want, with an aptitude. Expect to be a go-to person in it. Dont just  pick something to get an edge on a job, especially if you dont care for it. Yucky situation. 
 
2. Know thyself. If you are a go getter and confident, a split fellowship will work in most pups. The reality is you gain most your expertise on the job. A fellowship from 10 years ago has little meaning, its an iterative process. 
 
3. If you are a more nervous person, a year fellowship may be best, to really build confidence  I think it is mandatory for academics. 
 
4. I know nonfellowship mammographers who are excellent. Its about consciousness experience, especially with procedures- volume makes you better. Some fellowship rads tend to think the world revolves around breast imaging. They lose perspective, in part because for some of them it is their sole skill, so naturally from an ego perspective they overvalue their importance. Truth is, mammography is not intellectually challenging, it more about work ethic and experience. If motivated, a lot of people can do it well. Im not saying its easy, not at all. But it is a contained skill that can be attained with proper focus. I think a year fellowship and practicing Mammo 50% is the sweet spot. No matter what the fellowship, occasional rotations in Mammo is a mistake. You have to regularly exercise those muscles. 

5. minifellowships are a double edge sword during residency because they are at the expense of other experiences which are of value in most private practices. So I think a full residency with a split fellowship in two things you like is better. Dont cut too many corners, you have to put in the time to get the experience. Also, some fellowships require a year, particularly IR; no way around it imo. 

[Unknown Member]

1.  Do a fellowship in something you like and want, with an aptitude. Expect to be a go-to person in it. Dont just  pick something to get an edge on a job, especially if you dont care for it. Yucky situation.  
  
2. Know thyself. If you are a go getter and confident, a split fellowship will work in most PPs. The reality is you gain most your expertise on the job. A fellowship from 10 years ago has little meaning, its an iterative process.  
  
3. If you are a more nervous person, a year fellowship may be best, to really build confidence  I think it is mandatory for academics.  
  
4. I know nonfellowship mammographers who are excellent. Its about consciousness experience, especially with procedures- volume makes you better. Some fellowship rads tend to think the world revolves around breast imaging. They lose perspective, in part because for some of them it is their sole skill, so naturally from an ego perspective they overvalue their importance. Truth is, mammography is not intellectually challenging, it more about work ethic and experience. If motivated, a lot of people can do it well. Im not saying its easy, not at all. But it is a contained skill that can be attained with proper focus.
I think a year fellowship and practicing Mammo 50%in a practice is the sweet spot. More Mammo wont necessarily make you better.  No matter what the fellowship, occasional rotations in Mammo in a practice is a mistake. You are filling holes in a schedule, but not doing any favors towards patient care. Demand volume, or dont do it. The regulatory requirements are woefully low. You have to regularly exercise those muscles. Id rather have a full time extender interpret my wifes Mammo than an occasional radiologist interloper. 

5. minifellowships are a double edge sword during residency because they are at the expense of other experiences which are of value in most private practices. So I think a full residency with a split fellowship in two things you like is better. Dont cut too many corners, you have to put in the time to get the experience. Also, some fellowships require a year, particularly IR; no way around it imo.  

[spc44_369]

Is it ill advised to skip on fellowship completely if you have no interest in academics?

[william.wang_997]

Yes…ill advised…because if the current job doesn’t work out….the next one will need fellowship. All your other peers have fellowship, so if they are cutting; they might let non fellowship people go. With so many moonlighting opps during fellowship, not sure why you would skimp on that.
 
Let’s say you don’t do a fellowship as you wanted the extra dollars for 1 year, and then you have a kid and then a lifestyle…..5 years go by and there is downturn…and then you are scrambling as they are cutting back. 
 
 

[Unknown Member]

Not doing fellowship is a gamble. I agree about better to do it.

[dudapoll_jacomino]

“Six reader studies on digital mammography revealed a pooled sensitivity rate of 80.8 percent for stand-alone artificial intelligence (AI) in comparison to 72.4 percent for radiologist assessment while seven historic cohort studies showed a 75.8 percent pooled sensitivity rate for stand-alone AI versus 72.6 percent for radiologist interpretation of digital mammography.” From Diagnostic Imaging mag, reporting on [link=https://pubs.rsna.org/doi/10.1148/radiol.222639,]https://pubs.rsna.org/doi/10.1148/radiol.222639,[/link]
 
I’m not breast trained. I don’t mind mammo. The above study (there’s one like that every few weeks) is hopeful for taking away some of the drudgery of mammo screening work so we can do other things.
 
For PP and for a reasonably capable resident, 6 month breast fellowship is all I would recommend, preferably with another 6 months in body or msk, depending on what you like. 

[ruszja]

Quote from Trazobone

Is it ill advised to skip on fellowship completely if you have no interest in academics?

 
Yes. Many practices make it a point to only hire ‘fellowship trained radiologists’ as a tool for advertising. Doesn’t mean they require everyone to be fellowship trained in whatever they interpret, but its a required qualification.
The current market where the ability to fog a mirror and a radiology residency at some point in the past is all you need to get a job is an abnormality.

[DanielQuilli]

Breast imaging is simple not easy. There is a huge difference and if you understand that you will understand why relatively few rads truly like it.
 
Kind of like having a six pack, I could explain in less than 10 minutes how anyone could very simply have visible abs and a statuesque physique. Yet very few people can execute and youre more likely to be a millionaire than have visible abs. Its very simple but not at all easy.
 
Breast imaging is similar. I see all the time how rads (that coincidentally dont read mammo) talk about how “easy” mammo is yet I rarely see it being done well at the practice level.

[afazio.uk_887]

Mammo def not easy at all. 
 
Having a stable and regular patient base helps, as every year having a comparison is great. 

[tdetlie_105]

Quote from Waduh Dong

Mammo def not easy at all. 

Having a stable and regular patient base helps, as every year having a comparison is great. 

 
Diagnostics and complex biopsies/locs def challenging but RVU rich tomo-screeners not so much.  I had a screening day recently, pretty tress free compared to normal diagnostic days. Plus I produced about 20% more RVUs than usual.

[jonhanse_770]

Not to be that guy, but a 99.5% sensitivity means 5 out of 1k will be missed.

 
I never was a math major- apologies for the screw up It was my calculators fault…LOL
 
As crazy as it sounds most people are much more forgiving about a human missing something than a machine. People dont know or care about sensitivy and specificity- they expect the machine to be right, Promoting AI as a an adunct to a mammo diagnosis is the only way to go IMHO. Using it as the end all be all and the final decision is just an invitation to a bunch of problems.
 
I am sure many more than 925 cases are missed each year but that is the nature of the beast. Until an algorithm is developed that is 100% or we find a “superrad” who never misses a read misses will happen. Even at 100% someone will need to sign off on the case and make sure that what is read as a positive is indeed a positive in their skilled opinions. Negatives you just roll the dice and hope for the best,. At a minimum all postive cases identified by AI will need to be overread by a rad so that a false positive doesnt put a woman through any  unnecsesary procedures from a biopsy on down…
 
PACSMan

[cchandc]

Nuclear medicine is still around other than PET/CT and low volume thyroid therapy ?!

[smfst7_929]

AI wont replace anyone anytime soon, but breast imagers will be the first to be affected. Sure youll need people to do biopsy, but thats pretty much it. So the number of mammographers well need will be like 5% of current numbers.

[afazio.uk_887]

Its going to be Rads plus AI reading mammo. This will increase sensitivity and specificity of reads and is good for the patients.

Its like if you have every mammo double read by Rads today. Also would be more accurate.

But AI will need to show that it really is a valuable copilot and not just an overcaller / time waster.

My prediction is that it will be an overcaller for a long time.

[Unknown Member]

Right now screening is where the money is; so AI will help make screening way more productive and will be usurped by those that control the technology. So, professional revenue will be drawn away from rads, into the technical. Essentially the low hanging fruit will be stolen.
The labor intensive part, diagnostics, will remain in our domain. It will need subsidy, and mammographers will be beholden to someone for their jobs, unless they control the technical. Similar to IR today, it won’t be a profit center on the professional side. There will be a need of course, and therefore jobs, but the current golden days will be end.
Lesson: go into mammo because you dig it.  

[Unknown Member]

In my opinion. Generally speaking, breast imaging is a service that radiology departments – private, academics, or corporate – needs to staff to get the contract , and they need breast imagers on-site, like fluoro and IR, to do breast diagnostics including work ups and biopsies and locs. Thats in every hospital, big or small, in nearly every city or town across this great country. Thats a lot of places.

It doesnt matter if reimbursement becomes less than the CT and MRs that the group/hospital/company can have read from home or anywhere in the country that is the gravy train they can profit from, whether those profiting most are the radiologists or the hospital or shareholders, they need the breast radiologists on -site to get the contract and will pay them equal to other rads – or more – for that job, regardless of how much their work can bill for. Thats because its harder to find rads who will be on-site and move their family to live in that Wisconsin suburb, most rads want to work from home in the big city of their choice doing high volume work and not answer phones or talk to techs or consent patients, with some schedule that includes 7 to 14 days off at a time.

So I think its the other way around: rads who read MSK and Neuro and Body and arent needed on-site may be, one day, a dime a dozen and can be replaced more easily and have their compensation more easily decreased due to a larger and increasing supply of them, just look at how many people coming out of training (and later career) are considering remote Telerad jobs.

The willingness to do procedures and drive into work and show up on-site most days and move your family to where the job is located will keep breast imagers valuable in the face of AI, growth of corporate Telerad, and other commoditization.

[consuldreugenio]

Flounce is on point. Most mammo rads Ive met are divas. There are complaints galore from partners about their sweet set ups from an equal compensation with no call/weekends/holidays. But no one is volunteering to come on site to do diagnostic and needle work. Theres just something about breast imaging thats much different from the rest of diagnostic radiology. Most of it is easy, but your stomach with turn when you will no doubt encounter some of your screening and diagnostic misses.

Breast imaging will command big bucks for the next 15-20 years, just due to the need to be on site. If you do mammo and other things, you can likely make a very high salary like myself. You just have to program your mind to love it and say thanks for so much money.

[skysdad]

From a purely AI standpoint, breast imaging is in trouble. But the truly valuable ones who love their job and coming in to do locs all day aren’t going anywhere. 
 
So do you love breast or do you just love money?

[Unknown Member]

Just my 2 cents: I’ve done about 50/50 breast and other diagnostic work for over a decade, all breast is in hospital breast imaging centers or outpatient imaging center, and almost all non-breast diagnostic is now from home less than 4 miles from the hospital except when I do fluoro.  All of our groups’ breast imagers take call like the rest of the group members, taking equal call covering ER/inpt/urgent care on evenings and weekends and holidays.
 
At the end of the day, breast work – at least in my group, and I suspect elsewhere – is easier than non-breast work and that’s probably the consensus opinion in my group among those who do both, and half of our breast imagers did fellowships in something other than breast (neuro, body, pedi, etc).  Breast work is fairly algorithmic – like ordering from In N Out vs. Cheescake factory – and the technical skills are very limited, this isn’t IR; the last schedule patient is around 3:30pm or 4pm; and there is lots of banter all day long between rads and mammo/ultrasound techs, we have a blast. The breast rads routinely use the lunch break to take a walk together or go out to eat (I go to jiu jitsu class).  On any non-breast rotation, it is busy enough that even if you are able to leave the workstation for a lunch break or work out, you are strongly incentivized to keep grinding and stay abreast of the work. 
 
I think it’s great that non-breast imagers find breast so distasteful, part of me fears the day a couple of them (maybe the aging IR guys?) decide to do some breast and realize how easy it is compared to a bunch of ICU films, outpatient liver MRs, or thyroid ultrasounds. 

[afazio.uk_887]

I do general Rads in my gig but I find the breast work pretty pleasant and a positive aspect of the job. I know I am rare but thats good for job security.

[PraderWilli]

Following. What are people’s thoughts on the legal aspects and lawsuits in breast imaging?

[jonhanse_770]

     That is one of the main reasons why breast AI is slow to be accepted. If AI misses a call who do you sue?
     Insurers all want the lowst cost care possible, AI alone might provide this but again at what cost both financial and human? You may see AI do the primary read and a rad do a secondary one on breast screenings so they can have a throat to choke but if they are lucky rads will get 25% of what they are being paid now (if that) to do the screening. Diagnostic will almost always be rads first with AI assistance so hopefully there will be a CPT code that pays for AI’s involvement. It sucks.
 
Keep in mind you can quote all the sensitivity and specificity numbers you want but 99,.5% means that 50 cases out of ever 1000 will be missed. That doesnt include those misread by AI plus the time it takes to dictate why the rad says this and Ai that. Yes AI may catch a few more cases as well but at what cost? We learned this lesson with CAD a decade ago before CMS pulled the plug on reimbursement and the use of CAD went to hell in a handbasket,With so many vendors offering breast AI it wouldnt surprise me to find one who literally gives use of their software away and makes its money back from Tesla ads in between case reads…LOL.
 
Mike Cannavo “PACSMan”

[pankajkaira1982_700]

neurorad here. Used to do mammo but never became comfortable with it. Maybe I should do a mammo fellowship if it’s not as difficult as it seems. 
 
Some mental gymnastics required for stereotactic biopsy, no? 
 
 

[jonhanse_770]

Mammo is hugely in demand because few are comfortable with it, the risk to reward ratio imbalanced and the potential liabilities significant. You also dont need a fellowship but it sure helps witha bidding war sure to follow for contacting your services .
 
I recall working with a rad group several years back where the rad was interrupted (not that that never happens of course), and went back and dictated the study on memory without re-reviewing it. Very bad move. In the year from the mammo where the CA should been dictated and treatment initiated vs a normal interpretation (family history of breast CA- a clear as day diagnosis) when she had her follow up a year later the CA had metasticized and she was dead within the year..at age 36 …with 3 kids under age 10….Needless to say that was a very expensive open checkbook mess up (multi-million dollar uncontested settlement) not to mention the human costs of kids being without their mom…Needless to say the rad finding malpractice insurance also wasn’t easy either- and without it no one would accept them doing interpretations. Even with it doubts lingered about them moving forward with this person as a rad  since a settlement usually ends with the deepest pocket and no one wanted to be that… This rad ended up never reading another mamm again…ever…..even though it wasnt the interpretation that did them in but workflow-related issues. Easiest answer- phones on hold and a closed door with a sign on it that says “Reading in progress- please do not disturb” that is respected…Beleive it or not it can be done
 
Bottom line – It IS as diffucult as it seems…if not moreso…so IMHO stick with neuro…
 
PACSMan
 

[skysdad]

Quote from ThePACSman

     That is one of the main reasons why breast AI is slow to be accepted. If AI misses a call who do you sue?
   Insurers all want the lowst cost care possible, AI alone might provide this but again at what cost both financial and human? You may see AI do the primary read and a rad do a secondary one on breast screenings so they can have a throat to choke but if they are lucky rads will get 25% of what they are being paid now (if that) to do the screening. Diagnostic will almost always be rads first with AI assistance so hopefully there will be a CPT code that pays for AI’s involvement. It sucks.

[b]Keep in mind you can quote all the sensitivity and specificity numbers you want but 99,.5% means that 50 cases out of ever 1000 will be missed.[/b] That doesnt include those misread by AI plus the time it takes to dictate why the rad says this and Ai that. Yes AI may catch a few more cases as well but at what cost? We learned this lesson with CAD a decade ago before CMS pulled the plug on reimbursement and the use of CAD went to hell in a handbasket,With so many vendors offering breast AI it wouldnt surprise me to find one who literally gives use of their software away and makes its money back from Tesla ads in between case reads…LOL.

Mike Cannavo “PACSMan”

 
Not to be that guy, but a 99.5% sensitivity means 5 out of 1k will be missed.
Using some figures from google:
With a positive cancer rate of 0.5% it means 5 people out of 200k screened will have their cancer missed.
And if 37 million mammograms are done each year, that means only 925 women will have their cancer missed.
 

[Unknown Member]

Quote from sartoriusBIG

AI wont replace anyone anytime soon, but breast imagers will be the first to be affected. Sure youll need people to do biopsy, but thats pretty much it. So the number of mammographers well need will be like 5% of current numbers.

Breast surgeons can and will do all breast procedures. More RVUs for them.

[Unknown Member]

A graduate of the Breast Oncology training program should be able to care for all aspects of disease and/or provide comprehensive management. When referring to a discipline of training (e.g. research or community outreach), the graduate should be able to carry out the endeavor from its conception through completion.
 
2019 Breast Surgical Oncology Fellowship Curriculum and Minimum Training Requirements Effective August 1, 2019

[Unknown Member]

Should breast surgeons enter into mammography reading? One stop shop. Lucrative. Cardiology did it with their field.

[Unknown Member]

procedural skills are acquired during the breast surgical fellowship:
[ul][*]Interpretation of mammograms and ultrasound[*]Stereotactic or ultrasound-guided core biopsy[*]Subareolar duct excision[*]Intraoperative ultrasound[*]Breast-conserving surgery[*]Modified radical mastectomy[*]Skin-sparing mastectomy[*]Nipple-sparing mastectomy[*]Axillary dissection[*]Sentinel node biopsy [/ul] [link=https://college.mayo.edu/academics/residencies-and-fellowships/breast-surgical-oncology-fellowship-minnesota/curriculum/]https://college.mayo.edu/…-minnesota/curriculum/[/link]
 

[ruszja]

Do breast fellowship if you like the kind of practice and you want to devote a substantial part of your practice to breast. Don’t do it because through a quirk in the RVU methodology, screening mammography is currently overweight relative to other things we do.

It is one of the few areas of radiology where you can have a long term relationship with some patients and you can directly see the influence your care has on a patients life. We still have patients who ask to have Dr X read their screeners although he has retired many years ago.

[kaldridgewv2211]

So the concern is around AI and picking MSK or Breast.  If you’re worried about AI in mamms, wait till you hear about AZmed Rayvolve.

[jonhanse_770]

It is one of the few areas of radiology where you can have a long term relationship with some patients and you can directly see the influence your care has on a patients life. We still have patients who ask to have Dr X read their screeners although he has retired many years ago

 
I have a client who had a patient move out of state 10 hours away who made the road trip back to have the radiologist she had trusted for decades do her annual mamographic exam. She was not the only one either.
 
In mamography especialy, it’s all about trust.
 
 PACSMan