ObjectiveTo explore the impact of the time-of-flight (TOF) reconstruction on brain PET with short-lived 11C-labeled tracers in PET magnetic resonance (PET/MR) brain images among suspected patients with Alzheimer's and Parkinson's disease (AD/PD).MethodsPatients who underwent 11C-2-ß-carbomethoxy-3-b-(4-fluorophenyl) tropane (11C-CFT) and 2-(4-N-[11C] methylaminophenyl)-6-hydroxybenzothiazole (11C-PiB) PET/MRI were retrospectively included in the study. Each PET LIST mode data were reconstructed with and without the TOF reconstruction algorithm. Standard uptake values (SUVs) of Caudate Nucleus (CN), Putamen (PU), and Whole-brain (WB) were measured. TOF and non-TOF SUVs were assessed by using paired t-test. Standard formulas were applied to measure contrast, signal-to-noise ratio (SNR), and percentage relative average difference of SUVs (%RAD-SUVs).ResultsTotal 75 patients were included with the median age (years) and body mass index (BMI-kg/m2) of 60.2 ± 10.9 years and 23.9 ± 3.7 kg/m2 in 11C-CFT (n = 41) and 62.2 ± 6.8 years and 24.7 ± 2.9 kg/m2 in 11C-PiB (n = 34), respectively. Higher average SUVs and positive %RAD-SUVs were observed in CN and PU in TOF compared with non-TOF reconstructions for the two 11C-labeled radiotracers. Differences of SUVmean were significant (p < 0.05) in CN and PU for both 11C-labeled radiotracers. SUVmax was enhanced significantly in CN and PU for 11C-CFT and CN for 11C-PiB, but not in PU. Significant contrast enhancement was observed in PU for both 11C-labeled radiotracers, whereas SNR gain was significant in PU, only for 11C-PiB in TOF reconstruction.ConclusionTime-of-flight leads to a better signal vs. noise trade-off than non-TOF in 11C-labeled tracers between CN and PU, improving the SUVs, contrast, and SNR, which were valuable for reducing injected radiation dose. Improved timing resolution aided the rapid decay rate of short-lived 11C-labeled tracers, and it shortened the scan time, increasing the patient comfort, and reducing the motion artifact among patients with AD/PD. However, one should adopt the combined TOF algorithm with caution for the quantitative analysis because it has different effects on the SUVmax, contrast, and SNR of different brain regions.

Background The purpose of this study was to evaluate the effects of time-of-flight (TOF) reconstruction on regional brain positron emission tomography (PET) obtained from hybrid PET/magnetic resonance imaging (PET/MRI). Patients (n = 34) who underwent 18F-FDG PET/MRI brain imaging were retrospectively included in this study. The PET list mode data were reconstructed with and without a TOF reconstruction algorithm. Regional brain volumes of interest (VOIs, 18 regions) were segmented and their standardized uptake values (SUVs) were measured. Standard formulas were applied to measure contrast, signal to noise ratio (SNR), and percentage relative average difference of SUVs (%RAD-SUVs). The results were compared using paired t-tests. Results In all the VOIs, the SUVmax was significantly higher with TOF reconstruction than without it. The %RAD-SUVmax was greater than 3% for all VOIs in TOF compared to non-TOF reconstructions, and was largest in the caudate nuclei. The SUVmean values in 14/18 VOIs were significantly higher with TOF reconstruction, although the %RAD-SUVmean values were comparably smaller than the %RAD-SUVmax. The highest %RAD-SUVmean between TOF and non-TOF reconstruction was again in the caudate nuclei. Contrast and SNR were higher with TOF reconstruction in 12/17 and 9/17 VOIs, respectively. Although the differences were small, they almost showed statistical significance (p < 0.05). In the caudate nuclei, the percentage difference in SNR was nearly 50%. However, in the PA, SN, CoC, WM and CM, contrast was higher with non-TOF reconstruction, and SNR was higher in regions adjoining occipital lobe and nucleus accumbens. Conclusions With the TOF-technique, the SUVmax in the segmented VOIs was higher compared with non-TOF reconstruction, unlike SUVmean, which showed varying differences across the different VOIs according to the use of TOF and non-TOF reconstruction. The temporal lobe and caudate nucleus achieved considerably higher SNR with the TOF technique, but not the other regions. Overall, we recommend that these quantitative differences among VOIs are considered when diagnosing Alzheimer’s disease and Parkinson’s disease.

No abstract