An editorial in the Journal of Nuclear Medicine (J Nucl Med. 2007 Mar;48(3):331. PMID: 17332606) argues that combination PET/MRI systems are the future and will supplant PET/CT:
In a number of ways, the path to PET/MRI has been reverse of that to PET/CT. The first PET/CT design emerged from industry–academia collaboration and was a prototype for human clinical use that eventually stimulated a commercial response and led to the development of PET/CT for imaging small animals. In contrast, PET/MRI began with the small-animal design and then, over a decade later, the first PET/MRI brain images were acquired on a dedicated prototype system, following an impressive industrial backing that far exceeded that of the early PET/CT developments.
[…]
A mere 2 y after the advent of commercial PET/CT, Johannes Czernin from UCLA, at the 2003 annual DGN meeting, commented that “PET/CT is a technical evolution that has led to a medical revolution.” Today, at the dawn of PET/MRI, it may be said that “PET/MRI is a medical evolution based on a technical revolution.” Although PET/CT appears to have replaced stand-alone PET for most oncologic indications, it is reasonable to assume that PET/MRI will be the preferred imaging option for neurologic and central nervous system indications. Without doubt, such dual-modality combinations are here to stay because they incorporate the diagnostic power of PET. Thus, PET/CT and PET/MRI, by virtue of their combined anatometabolic imaging, will lead to a “new-clear” medicine and the demise of “unclear” medicine.
This piece has somewhat messianic feel to it. I remember looking for PET/MRI systems at Seattle ISMRM last year, there were only two talks as far as I recall (maybe someone can dig up the abstracts, I’ve lent out my CD and it never returned, grr). Is it really true that there is widespread industry commitment to making dual-mode systems a reality? The engeineering challenge seems an order of magnitude more complex than with the marriage of PET/CT. It would be a different matter if someone had built a prototype and was publishing extensively on the advantages, but the above implies that the manufacturers are making a leap of faith.
And I’m not sure either that of all the hybrid technologies that could be pursued, PET/MRI even makes sense. For one thing, SPECT seems to be making some inroads, and is not tied down to one radioisotope the way PET is. And of course the PET reimbursement issue is notoriously complicated, with no sign of improvement on the horizon. SPECT/CT hybrids are beginning to get more sophisticated, and it wouldn’t be hard to imagine a SPECT/MRI system. Or even a MRI/CT system, which would really be a money-saver from an outpatient imaging standpoint. If the hybrid has application to bread and butter imaging as well as exotic research and oncology applications, it’s going to do better on the market.
None of this is to say that no one should go for PET/MRI; it would be unquestionably cool and powerful. But is it really the inevitable slam-dunk as it’s being made out to be?
(full text of article available online)
Don’t PET imagers use photomultiplier tubes (vacuum tubes that use slow electrons)? I’d hate to be the engineer trying to make those work next to an MRI magnet.
I think that there are other forms of amplifiers that you can use. Really, a 7T magnet manages to throw a lot of wrenches around.
PET/MR just can’t replace CT. CT wins by so much on speed/resolution/geometric accuracy that it’s not even funny. Also, there isn’t a way to easily correct for attenuation in MR/PET like there is in PET/CT.
For that matter, if you have certain implants you can’t even walk into an MR suite.
Now MR/CT could technically replace PET, simply because MR can detect 19F, which a lot easier to get than 18F. The sensitivity might even be as good or better. This would require some major improvements with whole-body MR technology first, though.
PET/MR just can’t replace CT
note – I mistyped CT above when I should have typed PET/CT (now fixed). I see the overall point; PET/CT is a success because it marries anatomy (CT) with function (PET). MRI/PET would do the same, but offer functional imaging on both sides of the fence, and bring superior soft tissue contrast to the anatomic side. I agree Dustin that the attenuation issue is probably the single biggest obstacle (though one wonders, if PET/CT systems are the default nowadays anyway, why not merge MRI with PET/CT rather than just PET.. MRI/PET/CT!)
I think that PET is sexy but in the long term is too specialized. The economics of 18F dont lend themselves to flexible clinical imaging the way that SPECT can. The more I think about it, the more I believe that an MRI/SPECT combo would be way more useful overall.
Can MR/CT replace PET? Thats an interesting question. Do we even need CT to replace PET? What does CT really bring to the, er, table? Unless you mean, MRI/CT could replace PET/CT. But then if you’re gonna go through the ttrouble of integrating CT anyway, why not go for the trifecta system again 🙂
The combination of PET/MR/CT is the one most being looked at now. It is so difficult to produce that the design is becoming just one bed taking the patient through each scanner one by one. Some development of MRI/PET systems have been going on over at Phillips, but I am not sure what approach they are taking.
Concerning the photomultiplier tube problem, the two solutions are optical coupling and avalanche photodiodes. In optical coupling, the electromagnetic radiation is detected within the magnet and coupled optically to the PMT’s outside the magnet. Avalanche photodiodes are silicon devices that simulate the cascade effect of PMT’s but are virtually unaffected by high magnetic fields.
Also, I saw that you mentioned that ISMRM had very few talks on MRI/PET last year. Unfortunately, it looks to be the same this year. However, you should check out SNM this year. Most of the interest in MRI/PET systems will be there.