A Promising Future: Comparable Imaging Capability of MRI-Compatible Silicon Photomultiplier and Conventional Photosensor Preclinical PET Systems

We recently completed construction of a small-animal PET system-the MiniPET-3-that uses state-of-the-art silicon photomultiplier (SiPM) photosensors, making possible dual-modality imaging with MRI. In this article, we compare the MiniPET-3 with the MiniPET-2, a system with the same crystal geometry but conventional photomultiplier tubes (PMTs). Methods: The standard measurements proposed by the National Electrical Manufacturers Association NU 4 protocols were performed on both systems. These measurements included spatial resolution, system sensitivity, energy resolution, counting rate performance, scatter fraction, spillover ratio for air and water, recovery coefficient, and image uniformity. The energy windows were set to 350-650 keV on the MiniPET-2 and 360-662 keV on the MiniPET-3. Results: Spatial resolution was approximately 17% better on average for the MiniPET-3 than the MiniPET-2. The systems performed similarly in terms of peak absolute sensitivity (similar to 1.37%), spillover ratio for air (similar to 0.15), spillover ratio for water (similar to 0.25), and recovery coefficient (similar to 0.33, 0.59, 0.81, 0.89, and 0.94). Uniformity was 5.59% for the MiniPET-2 and 6.49% for the MiniPET-3. Minor differences were found in scatter fraction. With the ratlike phantom, the peak noise-equivalent counting rate was 14 kcps on the MiniPET-2 but 24 kcps on the MiniPET-3. However, with the mouselike phantom, these values were 55 and 91 kcps, respectively. The optimal coincidence time window was 6 ns for the MiniPET-2 and 8 ns for the MiniPET-3. Conclusion: Images obtained with the SiPM-based MiniPET-3 small-animal PET system are similar in quality to those obtained with the conventional PMT-based MiniPET-2.