Simon Cherry, professor and chair of biomedical engineering at UC Davis, discusses the combination of PET and MRI in a single scanner, which his lab built for laboratory mice studies, particularly in cancer research.Until now, what has kept MRI and PET scans from being combined?
The two modalities have a big potential to interfere with each other. On the PET side, you worry about the very strong magnetic field environment from the MRI, and on the MRI side, you worry about the eff ect the materials in the PET scan will have on the uniformity of the magnetic fi eld and also the possibility that the PET anatomics will generate radiation that will perturb the MRI image.
PET and MRI studies in the same subjects have been done for many years, but the subjects have been moved from one scanner to the other. It’s very common to use both modalities, just not at the same time.How did you overcome these issues?
In traditional PET scanners, the detectors are based on photomultiplier tubes, which are extremely sensitive to magnetic fields. We moved to what’s called an avalanche photodiode (APD), a silicon sensor that’s very thin. Because the active region is so thin, it’s much more immune to magnetic fi elds. We’ve tested these devices in magnetic fi elds up to 9.4 tesla, and they work just fine.
On the MRI side, a number of things are important. First, arranging the components of the PET scanner symmetrically in the MRI scanner helps to prevent preservation of the magnetic fi eld. We also carefully shielded the electronics of the PET scanner to try to prevent radiofrequency interference. And we used materials wherever possible that have reasonable magnetic susceptibility, so that they are similar to human tissues. It’s all careful engineering to help protect artifacts on the MRI side.What benefits will the combination provide to oncologists and their patients?
That’s the $1 million question for which we do not yet know the answer. We now have data that clearly show that this is feasible, and the question now is “are there applications out there that absolutely require PET and MRI data to be taken at the same time?” The things that spring immediately to mind are areas of the body in which MRI is a superior imaging technology to CT—the brain is an obvious case, as are the abdomen and pelvic areas. Particularly for oncology, you worry if you take the images a few hours apart in different scanners that things move around, and if you put the patient on two different beds, the organ positions are somewhat different.
It’s very hard to use computer software to register those two data sets. So, the ability to take the PET and MRI at the same time and know exactly where the PET’s localized with respect to the anatomy you get from the MRI, those applications are key. Just like PET/CT has had a huge role in oncology, I think PET/MRI will be the same. I think there is going to be a period of using this in the research arena to really understand what unique things can be done. For it to be a common, mainstream diagnostic tool, many other factors come into play, economics of course being one of them.That’s harder to predict: will there be an application for which it makes a significant diagnostic difference and for which you can justify the cost (this is not going to be cheap)?
If you take this whole emerging field of molecular imaging, in which we’re really trying to image for example, specific sets of enzymes, they’re developing targeted agents to do this. As a tool that lets you really understand what you’re looking at and where the signal is coming from, the PET/MRI combo will be a really great device. For validating new biomarkers—many people are interested in using MRI as a biomarker for therapeutic effect—and to validate that specific radio tracing with PET in a pre-clinical situation and then clinical trials will be very helpful. I see a lot of applications in the research and drug development area. But my concern at a diagnostic level, as I mentioned, is where are the applications that justify the cost in a healthcare economy where, of course, costs are being tightly controlled for good reason. That’s going to be the challenge, but everyone said the same thing about PET/ CT 10 years ago; they weren’t quite sure what it was going to be used for. There were doubts about it being costeff ective. Now it is a mainstream modality, and nobody buys a PET scanner without a CT attached anymore. Maybe it will be the same with PET/MRI.
We’re trying now to develop a much-higher performance, second-generation system, and we’re also investing a lot of effort in trying to develop new imaging probes that are both MRI- and PET-visible at the same time.