The technology presented in the paper is pretty bleeding edge and represents 2 years of instrumentation and development work. They present a prototype PET unit for use in a small animal MR magnet operating at 7T. Their solution to solving the problem of detecting light from the scintillator crystals was to use position sensitive photodiodes, which are less prone to distortions from magnetic field effects. In order to reduce electrical noise in the detectors, the authors used a cold nitrogen gas to cool the detectors. This is understandably a significant limitation for real world clinical work, but not insurmountable. Several interesting effects were noted with using the photodiodes and fiberoptic coupling and are nicely illustrated with sample images.
Some very interesting development work here that shows a lot of potential. It's not the first PET/MRI hybrid unit developed or the only one being worked on, but the design and implementation the authors have come up with has the potential of retrofitting existing MR scanners with the capability rather than having to get a new magnet. Obviously it's still several years away from any kind of implementation for human use, but in the meantime a working unit for small animal imaging would probably yield some very useful information for researchers.
Ciprian Catana, Yibao Wu, Martin S. Judenhofer, Jinyi Qi, Bernd J. Pichler and Simon R. Cherry, "Simultaneous Acquisition of Multislice PET and MR Images: Initial Results with a MR-Compatible PET Scanner", J Nucl Med 47: 1968-1976
PET and MRI are powerful imaging techniques that are largely complementary in the information they provide. We have designed and built a MR-compatible PET scanner based on avalanche photodiode technology that allows simultaneous acquisition of PET and MR images in small animals. Methods: The PET scanner insert uses magnetic field-insensitive, position-sensitive avalanche photodiode (PSAPD) detectors coupled, via short lengths of optical fibers, to arrays of lutetium oxyorthosilicate (LSO) scintillator crystals. The optical fibers are used to minimize electromagnetic interference between the radiofrequency and gradient coils and the PET detector system. The PET detector module components and the complete PET insert assembly are described. PET data were acquired with and without MR sequences running, and detector flood histograms were compared with the ones generated from the data acquired outside the magnet. A uniform MR phantom was also imaged to assess the effect of the PET detector on the MR data acquisition. Simultaneous PET and MRI studies of a mouse were performed ex vivo. Results: PSAPDs can be successfully used to read out large numbers of scintillator crystals coupled through optical fibers with acceptable performance in terms of energy and timing resolution and crystal identification. The PSAPD-LSO detector performs well in the 7-T magnet, and no visible artifacts are detected in the MR images using standard pulse sequences. Conclusion: The first images from the complete system have been successfully acquired and reconstructed, demonstrating that simultaneous PET and MRI studies are feasible and opening up interesting possibilities for dual-modality molecular imaging studies.