67 E ect of flow on longitudinal tau PET For AD subjects each (100min) scanning session consisted of two dynamic PET scans of 30 and 20 minutes respectively, with a 50 minute break in between[8]. Details are described elsewhere[8]. In short, the first 30 minute dynamic scan started simultaneously with a bolus injection 236±11 MBq 18F-flortaucipir (injected mass 1.1±0.8 µg). The second PET scan was co-registered to the first dynamic PET scan and PET list mode data were rebinned into a total of 23 frames (1x15, 3x5, 3x10, 4x60, 2x150, 2x300, 4x600 and 4x300 seconds). All data was reconstructed using 3D RAMLA with a matrix size of 128x128x90 and a final voxel size of 2x2x2 mm3, including corrections for dead time, decay, attenuation, randoms and scatter. T1-weighted MR scans were co-registered onto the corresponding PET images using Vinci software. Volumes of interest based on the Hammers[9] and Svarer [10] template were subsequently delineated on the MR images and superimposed on the PET scan using PVElab[10]. Corresponding regional time activity curves (TACs) were extracted by superimposing these regions of interest from both templates onto the dynamic PET scan. Binding potential (BPND) and R1 parametric images were generated using receptor parametric mapping (RPM) with cerebellum gray matter as a reference region[11]. The gap of 20 minutes in the reference region TAC was interpolated by using cubic interpolation for the SCD subjects[11]. For the AD patients, population based plasma-input-function in combination with a reversible two tissue compartmental model with blood volume correction (POP-IP_2T4k_VB) was used to interpolate the gap of 50 minutes in the reference tissue TAC [8]. Methods: Reference region check The choice of reference region for 18F-flortaucipir is challenging, especially in a longitudinal setting. In the current study, we used cerebellum gray matter and since there was no gold standard available (e.g. plasma-input data) to validate the reference region in the present dataset, we utilized a quality check method which was published by de Vries et al. [12] . We compared the non-invasive standardized uptake values corrected for lean body mass (SUL80-100min) for each cerebellum gray matter of baseline and follow-up of both SCD subjects and AD patients and we observed that the cerebellum gray matter SUL80-100min remained similar between different subject groups and time points. Methods: Simulations Simulated TACs were generated to assess the effect of (larger) flow changes on the accuracy of DVR/BPND and SUVr values. Non-linear regression (NLR) based simplified reference tissue model (SRTM) BPND and SUVr derived from the clinical cohort were used for this end. Our group previously validated the use of SRTM for 18F-flortaucipir 3
RkJQdWJsaXNoZXIy MjY0ODMw