3.2.2. SiPM Technology

The increase in effective sensitivity enabled by modern SiPM could be translated into an equivalent reduction in injected activity or acquisition time while retaining equal image quality as PMT [49,51]. Intraindividual comparisons of clinical PET data between scanners equipped with either SiPM or PMT are scarce because this requires a second scan in a randomized protocol to prevent bias from systematically different uptake times between the scans. Sekine et al. used a randomized protocol to investigate the potential

to reduce injected [18F]FDG activity required (or acquisition time) through the use of an SiPM TOF PET/MRI instead of a PMT TOF PET/CT in 74 patients with different types of malignant tumors. Image quality (artifacts, noise, and sharpness) was rated as acceptable at up to 40% reduction in simulated acquisition time with the SiPM PET/MRI. However, the potential to reduce acquisition time with the PMT PET/CT was not specifically investigated. Moreover, the SiPM PET/MRI had a 25 cm axial FOV compared to the 15.7 cm FOV of the PET/CT system, which partly explains differences in image quality (especially image noise) independent of the photomultiplier technology [153].

<sup>L</sup>ópez-Mora et al. reported higher lesion detection rates with a SiPM PET/CT compared to a PMT PET/CT in 22 of 100 patients using [18F]FDG or [18F]fluorocholine (58 patients underwent SiPM PET first; axial FOV were similar). This resulted in a modified disease stage in 7 of these 22 patients based on SiPM PET/CT [154]. Similarly, Baratto et al. found higher lesion detection rates in 13 of 94 patients with [68Ga]Ga-DOTATATE using SiPM vs. PMT PET/CT in randomized order (SiPM, 20 cm axial FOV; PMT, 15.7 cm) [155].

Consequently, SiPM technology may well produce improvements in image quality and lesion detection rates. However, the magnitude of improvement that can be achieved in clinical scans is likely to vary with different SiPM designs.
