Impact of Time-of-Flight Reconstructions on Maximum Standard Uptake Value in PET/MRI: Using of 18F-FDG, 11C-CFT and 11C-PIB

Abstract Objective: To evaluate the impact of time-of-flight (TOF) reconstruction on the maximum standard uptake value (SUV max) among 18F-FDG, 11C-CFT and 11C-PIB in positron emission tomography/magnetic resonance imaging (PET/MR) brain images.Methods: Suspected Alzheimer’s and Parkinson’s patients (AD/PD) who underwent 18F-FDG, 11C-CFT or 11C-PIB PET-MR imaging were retrospectively included in the study. Each PET LIST mode data was reconstructed with and without TOF respectively using the system default three-dimensional ordered subset expectation maximization iterative reconstruction algorithm (3D-OESM). SUVmax of the selected brain volume of interest (VOI) were measured using the PMOD 3.906 PNEURO module in both TOF and non-TOF images. TOF and non-TOF SUVmax values were assessed by paired t-test and Pearson’s correlation graphs. Contrast improvement of VOIs, relative % SUVmax difference and the percentage signal-to-noise ratio (%SNR) of whole-brain were measured using standard formulas.Results: A total 109 patients were included with the median age (years) and body mass index (BMI-kg/m2) of 62.2±6.8 y and 24.7±2.9 kg/m2 in 11C-PIB (n=34), 60.2±10.9 y and 23.9±3.7 kg/m2 in 11C-CFT (n=41) and 62.3±9.0 y, 23.6±3.5 kg/m2 in 18F-FDG (n=34), respectively. Higher average SUVmax values were observed in all the brain VOIs in TOF compared to non-TOF reconstructions, and relative % average SUVmax difference for all segmented brain VOIs were positive for all three radiotracers. Differences of SUVmax were significant in 100% of segmented VOIs in 18F-FDG (P<0.01), similarly 83.34% of VOIs in 11C-PIB (P<0.05), inversely not in 11C-CFT. Contrast improvement was showed significant (P<0.05) for 60.0%, 66.7% and 50.0% of segmented VOIs in 18F-FDG, 11C-PIB and 11C-CFT in TOF, respectively. %SNR gain for whole-brain showed slightly higher in 11C-PIB and 18F-FDG only in TOF, inequitably difference was significant (P<0.05) in 11C-CFT non-TOF images. TOF versus non-TOF reconstructions were strongly correlated positively and differences were significant (P<0.01) in all the VOIs among three radiotracers. Conclusion: The TOF reconstructions significantly improve the SUVmax value among all the brain VOIs in 11C-PIB, 11C-CFT and 18F-FDG. Consequently, it enhanced image contrast and SNR. Thus, the quantitative effect of TOF reconstruction in simultaneous PET/MRI may positively impact the proper diagnosis of some diseases, such as Alzheimer’s and Parkinson’s diseases.