HN user

riahi

518 karma

Assistant Professor in Radiology and Nuclear Medicine.

My academic interest is applications, validation, and deployment of machine learning to clinical radiology in the context of Radiology/Operations/IT deployments.

MD/Residency/Fellowship from Harvard Medical School.

Posts3
Comments214
View on HN

You can see it in just this PDF report.

It's multiple things. It never shows the subscapularis in the way that people actually look the tendon. It hyper fixates on the axial when I find the sagittal much more useful for subscapularis.

Figure 7. There's an arrow pointing "to the acromial undersurface". The arrow is not pointed to that location.

Figure 5. "thin bursal fluid". This is within physiologic variation, but is calling bursitis.

It keeps bringing up irrelevant normal things like the shape of the coracromipal arch, I assume because lots of websites have information about that as a patient focused possible cause for rotator cuff impingement.

I am reminded of the recent Stanford MIRAGE study which found that LLMs will happily hallucinate answers about medical images if the medical images are omitted.

https://arxiv.org/html/2603.21687v2

This data is from 2006. Over 20 years, there has been substantial progress in CT radiation reduction through model-based iterative reconstruction and now ML-assisted reconstruction, aside from iterative advances in detector sensitivity and now photon-counting CT.

In clinical practice, those doses are about 2-3x what I see on the machine dose reports every day at my place of work.

In thin patients who can hold still, I've done full-cycle cardiac CT and achieved a < 1 mSv dose. We are always trying to get the dose down while still being diagnostic.

Source: Practicing radiologist.

I had no shop class in my middle school years 99-2002 but technical theater is the closest thing that may still sneakily exist in many schools. I learned basic carpentry and electrical, including lighting design and sound, all in the service of making sets for the yearly plays.

RE horizontal - there is a ribbon cable that can literally fracture which causes the Zr and Zl buttons to quit working which only really manifests when trying to use 1 joycon horizontally (personally when Mario party happens).

The repair takes about 20 minutes the first time you do it and the ribbon cable is on amazon for about $7.

Strictly speaking, phenylphrine works when it’s injected. It is approved and used every day for this indication in the service of general anesthesia.

However, the oral bioavailability is zero.

This is an example of using something off label that’s approved for something else. Sometimes it’s fine. And sometimes, it’s dumb.

Nit: isn’t “playing” itself “nouning the verb” via a gerund / participle construct?

I actually am not sure of the subtleties on this grammatical point.

It’s because oral minoxidil requires a ramp period and some people can become symptomatic. If you browse Reddit or other forum threads about it, look for a discussion on “sides”.

There are several papers in the dermatology literature suggesting low dose oral minoxidil for hair loss. I think dermatologists or hair surgeons who do this every day will know the literature. The average pcp likely won’t unless asked.

Agree. Also an MD.

My own personal experience with deceased donor transplantation is the pre-donor screening imaging to make sure there’s not a surprise cancer hiding for the recipient…

This had the flavor of being written by one of those types who’d rather write breathless blog posts about taking care of patients than actually taking care of them.

If you’ve been to medical school, you know the type.

For a pure cognitive specialty, sure. But watch out if they want to do any procedures outside of a hospital. The capital / regulatory heavy specialties are getting PE-roll-upped and that’s the same story as Veterinarians.

The new world is that we are becoming bound to hospitals / private equity in a way akin to sharecroppers on a plantation, mostly because of regulatory uncertainty which requires tremendous capital to overcome.

Some states are making it easier for surgeons to open surgicenters, but those are still very tightly regulated. Same with freestanding ERs, birthing centers, radiation therapy treatment centers, or freestanding imaging centers. Even if I wanted to raise the capital myself, the regulatory uncertainty makes the playing ground so unfair that it is functionally impossible without bringing in PE money to fund the legal fight for approval.

——— I trawled the approvals this year in my state.

Hospital joint venture with PE-backed imaging chain gets approval to open new freestanding imaging centers. They have the funds to fight back against the other systems’ legal challenges.

Radiologist owned 20 year old magnet who literally charges cash price of $499 (compared to above hospital planning to charge $1911 vs $6000 in hospital) was denied permission to upgrade to a newer magnet.

Physician practice request to buy a PETCT blocked by a competing hospital for 2 years of litigation, now referred to state Supreme Court…

——————- It doesn’t make sense that I can open a medispa and offer laser dermabrasion and other cosmetic services easier than I can open an imaging center. My options are join a hospital or join PE (rad partners or USRS).

Honestly I should move.

I’ll actually push back against this narrative.

A large problem is that people are coming out with insurmountable debt burdens from professional school and are simply unable to balance that and the start up costs of a new practice.

Combine that with states having adopted certificate of need requirements since the late 1980s, it is literally illegal for some doctors to even try to open their own practice.

In my state, I legally cannot open a radiology imaging center without spending thousands of dollars on a petition to the department of community health begging “please approve my imaging center” which then is posted publicly on a tracker website where in-house counsel for every single health system / competitor can sue to block you from ever opening.

They can also sue to block you from upgrading your equipment when it’s long out of date.

It’s untenable. While I agree there are many problems with poor financial savvy amongst highly trained professionals, I also realize that the average debt burden is over 250k and practice start up costs are not getting cheaper.

A PET or MR or CT could easily cost 500k-1million alone, without including real estate costs, building/renovation costs, or bribes/legal fees battling for the right to open your doors.

They just call out the surgery in an attempt to claim that with a tubal ligation, talc should be blocked from moving up the fallopian tubes. They don't even realize they contradicted their own results.

They did address this and noted that in patients with tubal ligation the talc distribution was different and lymphatic rather than “mechanical” through the fallopian tubes.

Still not sure what your angle is here.

It’s known that talc is inflammatory. It’s literally used with that intention medically. We don’t use powdered latex gloves in the modern day nearly as frequently as before, mostly because of latex allergy but also due to concerns that talc exposure leads to worse peritoneal scarring/adhesions.

The fact that talc is present inside the surgical specimens is bad in and of itself, regardless of how it got there. That implies an environmental cause for the malignancy, a la asbestos or smoking.

The peritoneal cavity and the pleural cavity have very similar linings.

Talc introduced into the peritoneum is similarly inflammatory and has been known since at least 1943:

https://jamanetwork.com/journals/jama/article-abstract/26104...

Work has been done to demonstrate that talc can migrate from external perineal application into the female reproductive tract.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6779257/

Talc is used medically as a treatment for recurrent pleural effusions. It is introduced into the pleural space and triggers an intensely inflammatory reaction which more or less fuses the pleura together to prevent the accumulation of fluid (pleurodesis).

When a patient who has had a pleurodesis gets an FDG PETCT, they will have intense accumulation of radiotracer for years. It is very inflammatory. Even more so than asbestos.

I wouldn’t want talc anywhere inside my reproductive tract or anywhere inside my body if I could avoid it.

If you buy the chronic inflammation leads to cancer hypothesis, talc leading to cancer is not so far fetched.

Sometimes we have the clinical context, usually if practicing in a large hospital system with an integrated EMR. It's not usually so neatly summarized though; maybe if we are lucky we can quickly glance through relevant notes at the time of scan interpretation.

However, healthcare in the US is very fragmented. Many patients seek cheaper imaging at freestanding imaging centers. Those places often don't have the same HIT integrations to have similar medical context.

And in those settings, I only know what's on the images and maybe 200-300 characters on the "reason for study" box.

This is not to say I think everyone should get scanned at expensive sites; more an indictment on how annoying the current EMR situation is.

Tough case. Adrenal adenomas are common incidental findings. ~15% are hormonally active, which means the vast majority are not.

As a radiologist, I sometimes wonder about whether I make too many recommendations to referring doctors (consider endocrine evaluation for a potentially hormonally active adrenal nodule).

A FREQUENT attack on us as a specialty is that we "find too many incidentals" (see attacks on mammography, breast cancer screening, other sorts of screening, ad nauseam).

Perhaps I'll keep doing the adrenal nodule recommendation, although I usually only make the recommendation if it's 1cm or larger.

I don’t think it’s heating. I think it’s PNS because my leg will periodically twitch which I can time with the pulse sequence.

I also have some small ACL reconstruction hardware in the knee which might interact and predispose that leg to moving…

I am well acquainted with high SAR abdominal scans in a 3T from my time doing scan certification in a petmr. Even though I “know” that tissue heating from high SAR is a thing, it always surprises me when my abdomen gets warm.

The RF pulses and field gradients can directly stimulate the peripheral nerves. I have had involuntary muscle twitching triggered by many pulse sequences. It’s not unheard of. It made me more sympathetic for patients who have motion in images. It could be that the particular combination of pulse sequence and their nerves don’t jive and triggers movement.

Source: Radiologist (and personal experience in the bore)