B How much should it cost to build an MRI Scanner?

[Myles Cummings]

I recently had an MRI scan, and while I was inside I began thinking about how the thing actually worked, and what it would cost to build. The hospital staff told me that the cost was several million dollars, and subsequent research seems to confirm that the 'average' MRI scanner cost between 3-5 million dollars.

I wondered whether it may be possible to build one for less?
After all in principle isn't it is just a very large set of electromagnets, a few radio antennas and a cooling system? OK I understand its a bit more sophisticated than that and it has to be computer controlled etc within a specially constructed room, but I think you get my drift?

Would it be possible to construct a viable MRI scanner for less - say sub 1m dollars?

Interested to hear your opinions, and apologies if this isn't the right thread to post this question, it's my first time here.

Cheers

M

 

[Dale]

Would it be possible to construct a viable MRI scanner for less - say sub 1m dollars?

Yes. The major manufacturers do offer such systems. For instance the Siemens Magnetom Essenza is a 1.5 T superconducting system priced at < 1 M USD. Refurbished and/or low field systems can be obtained for < 0.5 M USD.

 

[Myles Cummings]

Hmm. Interesting. Thanks for the input.

 

[Dale]

Hmm. Interesting. Thanks for the input.

Welcome! You might think about this question: if such low cost MRI systems exist, then why is a typical system so much more expensive?

 

[russ_watters]

For instance the Siemens Magnetom Essenza is a 1.5 T superconducting system priced at < 1 M USD.

Also note that that's probably twice the actual cost to build it. There's shipping, installation and start-up, warranty/service, and of course overhead and profit in there.

 

[Tom.G]

For small items at least, roughly pocket radio complexity, the minimum multiplier for selling price is 5 times material cost. That covers the minimal mark-up in the supply chain, without advertising, field support etc. Much stuff on the market is at least a 10 times multiplier.

 

[Ian Cumming]

How much to get all the image processing hardware and software to work successfully? It did take a few ( well, more than 20 ) PhDs in various fields to put it together in the first place, the imaging was the bottle neck. That is the break through that Cho Zang-hee pursued at Cololombia/KAIST in developing it. The strength of the magnet is also a large part of the production. It has to maintain a very uniform field inside the doughnut to get good images.

Keep in mind also that GE General Electric gave up on MRI/NMR saying it was unusable.

After you have a great working knowledge of the science, ( I would spend a few years studying all the papers listed...) You could possibly build your own...

I have a connection to this: My father was Cumming, J.

https://en.wikipedia.org/wiki/Cho_Zang-hee