An alternative to post hoc image fusion by software is, instead, to fuse the hardware from the two imaging modalities. While presenting a significant number of challenges, such an approach overcomes many of the difficulties of the software fusion methods.
In the early nineties, Hasegawa and co-workers at the University of San Francisco developed the first device that could acquire both, anatomical (CT) and functional (SPECT) images, using a single, high-purity germanium detector for both modalities [10, 11]. The CT images were, in addition, used to provide attenuation factors for correction of the SPECT data , and operating the device with two different energy windows allowed simultaneous emission-transmission acquisitions to be performed. This pioneering work of Hasegawa, Lang and co-workers is important because it was one of the first to take an alternative, hardware-based approach to image fusion. However, the difficulty of achieving an adequate level of performance for both SPECT and CT with the same detector material and without compromising either modality, led the group to explore a combination of SPECT and CT using different, dedicated imaging systems for each modality -a clinical SPECT camera in tandem with a clinical CT scanner . The CT images are used to correct the SPECT data for photon attenuation, and the device has been used clinically for patient studies since around 1996.
Recognizing the advantages of a hardware solution to combining anatomy and function, Townsend, Nutt and co-workers initiated a design for a combined PET and CT scanner in the early 1990s. The device comprised a clinical CT scanner and a dedicated clinical PET scanner. As with the CT/SPECT scanner of Hasegawa, the CT images are also used to correct the emission data for photon attenuation. The device was completed in early 1998 and underwent an extensive, three-year clinical evaluation programme at the University of Pittsburgh.
A third possibility to have anatomy and function imaged in a single device is to combine PET and MR. Obviously such a combination is technologically more challenging than combining PET with CT in view of the extensive restrictions placed on the imaging environment by the strong magnetic field. Nevertheless, proposals to place PET detectors inside an MR scanner also date back to the mid-nineties [14,15]. In 1996, an MR-compatible PET scanner was developed at UCLA , and then in 1997 a second, larger prototype (5.6 cm diameter ring) was constructed  and used in collaboration with researchers at Kings' College London for phantom and animal studies. The studies clearly demonstrated that simultaneous PET and MR images could be acquired using a range of pulse sequences and at different field strengths . The device opened up the possibility of simultaneously imaging [18F]-FDG uptake and measuring MR spectra. Currently a larger, 11.2 cm PET detector ring is being developed, designed to fit inside a 20 cm diameter magnet bore . However, scaling the design up to human dimensions will present many challenges and is still a number of years away.
Was this article helpful?