Quantitative and Visual Assessments toward Potential Sub-mSv or Ultrafast FDG PET Using High-Sensitivity TOF PET in PET/MRI

Newer high-performance time-of-flight (TOF) positron emission tomography (PET) systems have the capability to preserve diagnostic image quality with low count density, while maintaining a high raw photon detection sensitivity that would allow for a reduction in injected dose or rapid data acquisition. To assess this, we performed quantitative and visual assessments of the PET images acquired using a highly sensitive (23.3 cps/kBq) large field of view (25-cm axial) silicon photomultiplier (SiPM)-based TOF PET (400-ps timing resolution) integrated with 3 T-MRI in comparison to PET images acquired on non-TOF PET/x-ray computed tomography (CT) systems. Whole-body 2-deoxy-2-[F-18]fluoro-D-glucose ([F-18]FDG) PET/CT was acquired for 15 patients followed by whole body PET/magnetic resonance imaging (MRI) with an average injected dose of 325 +/- 84 MBq. The PET list mode data from PET/MRI were reconstructed using full datasets (4 min/bed) and reduced datasets (2, 1, 0.5, and 0.25 min/bed). Qualitative assessment between PET/CT and PET/MR images were made. A Likert-type scale between 1 and 5, 1 for non-diagnostic, 3 equivalent to PET/CT, and 5 superior quality, was used. Maximum and mean standardized uptake values (SUVmax and SUVmean) of normal tissues and lesions detected were measured and compared. Mean visual assessment scores were 3.54 +/- 0.32, 3.62 +/- 0.38, and 3.69 +/- 0.35 for the brain and 3.05 +/- 0.49, 3.71 +/- 0.45, and 4.14 +/- 0.44 for the whole-body maximum intensity projections (MIPs) for 1, 2, and 4 min/bed PET/MR images, respectively. The SUVmean values for normal tissues were lower and statistically significant for images acquired at 4, 2, 1, 0.5, and 0.25 min/bed on the PET/MR, with values of - 18 +/- 28 % (p < 0.001), - 16 +/- 29 % (p = 0.001), - 16 +/- 31 % (p = 0.002), - 14 +/- 35 % (p < 0.001), and - 13 +/- 34 % (p = 0.002), respectively. SUVmax and SUVpeak values of all lesions were higher and statistically significant (p < 0.05) for 4, 2, 1, 0.50, and 0.25 min/bed PET/MR datasets. High-sensitivity TOF PET showed comparable but still better visual image quality even at a much reduced activity in comparison to lower-sensitivity non-TOF PET. Our data translates to a seven times reduction in either injection dose for the same time or total scan time for the same injected dose. This "ultra-sensitivity" PET system provides a path to clinically acceptable extremely low-dose FDG PET studies (e.g., sub 1 mCi injection or sub-mSv effective dose) or PET studies as short as 1 min/bed (e.g., 6 min of total scan time) to cover whole body without compromising diagnostic performance.