Is the capital cost of the machine the expensive part of an MRI? The cost of many medical procedures is more about the staff than the equipment.

True, but this isn't a medical procedure. There are no surgeons etc involved. It amounts to the time for one mri tech to put the person in the machine, push the button. There is some expense in the reading and interpretation of the scans, but it's nowhere near the labor and other costs of surgury or endoscopy.

> this isn't a medical procedure

MRI techs and radiologists aren’t cheap.

Perhaps AI can optimize MRI operating costs. But the question then is whether those savings would be passed on to the patient or accrued by the manufacturer and hospital as extra margin.

We all know how that's going to play out

Only needed 1 tech for my MRI which lasted well under an hour. They may be expensive, but they don't get $100 an hour, so hard to believe that is a cost driver for a scan that costs over $1000.

1 tech to take the scan. And another one to actually examine/interpret the scan. Which for a full-body scan with no symptoms or specific areas to examine takes a lot longer than just taking the scan.

> They may be expensive, but they don't get $100 an hour

Radiology is one of the highest paying medical specialties. My cousin is one of the radiologists who interprets scans. He makes about $500k per year. Searching online, about $170/hr is average.

https://www.ziprecruiter.com/Salaries/Radiologist-Salary-per...

https://www.salary.com/research/salary/alternate/radiologist...

Yeah but a scan takes one radiology tech 30 mins, then the radiologist does the analysis, again not hours.

In surgical procedures all those people have to physically be there for hours

Yes, a properly done full body scan interpretation will take hours. It's about 30 minutes to analyze a scan of just someone's kidneys when there's already symptoms and doctor's note on what to look for.

The longest part is often just the paperwork and noting what was not seen.

The radiologist is a specialized medical doctor. They are not cheap. My neighbor is one, and she makes a considerable amount of money. They’re not technicians, and good ones command more than most surgeons.

Yes, but they can sit at a screen and analyze this stuff all day. It's not like a surgury where you have 3 or 4 OR staff and dozens of support staff to account for who all have to be on site and scrubbed in, etc.

There's a reason the body scan is $2500 and a simple surgery with anaesthesia starts at like $15000 before insurance.

Exactly the point I've been trying to make, thank you

That still doesn't mean radiologists aren't the main cost of a scan. Just that it doesn't cost as much as a team of 10.

True, was never trying to make the first point

That's the point of the comment thread were in. If you weren't trying to make that point, I don't get the relevance to the thread.

MRIs use helium, which ain't cheap, although truthfully I don't know if the helium runs dry fast or if you can use the same helium supply indefinitely.

The helium is used whether you use the machine or not, it is just keeping the superconducting magnet cool (and you can't really put it out of field without an incredible cost and risk of never being able to put it in field again). Plus most recent machines have decent recycling systems (that do not reduce to 0 but over 90% reduction in many cases)

The helium is one of the lowest cost parts of an MRI machine.

There is a big heavy spinning magnet that spins around your whole body at like 50 rpm. It's not an easy thing to make.

MRI machines don't spin. You are thinking of CT scanners.

You are not talking about an MRI, nor a CT.

Magnet is fixed in an MRI and you just pulse fields in the XYZ directions to create gradients of magnetic field. The pulse emission and receiving antennas are fixed too.

In a CT scan, what is rotating is one or many xray sources and one or many detectors plus some electronics and cooling systems if needed.

PET scans have fixed rings of pairs of detectors. But they can also be combined with a CT in the same instrument.

Thanks, my bad.

Revising my point to be that phasing those fields and controlling the power, is expensive and complicated

All of that is included in the initial price of the machine. There is no real cost associated when running the machine itself (outside of when a preamp or pulse generator burns but that's rare).

What’s the turn around time for a full body scan these days anyways? 1hr? More ? Less ?

I think the bottleneck is that a doctor has to look at it.

And when you don't have any particular symptoms it is hard to put as much effort into checking every millimeter of the body

We had advanced image processing techniques to detect tumors almost 20 years ago. With more advancements in computing power and AI techniques, it should be so much quicker and easier to do the same. The AI need not be conclusive but it could be a great assist to a radiologist who is looking at the scans.

To be clear, the coolant system runs continuously, but the power draw of the machine is far lower when not being actively used.

why do they have to run continuously? I would expect the opposite to be true. It would have some cool-off periods etc.

MRIs use superconducting coils which can never be shutdown except in emergencies. There's a "quench" button which if you press the magnet turns off. But then I can't remember how much it is to turn back on but it's not a simple visit from a mechanic.

Considering they’re superconductors cooled by liquid helium there’s not a lot of scope for “cool-off” lol