Fig. 2: Total-body positron emission tomography (PET)-enabled high-temporal resolution (HTR) dynamic imaging and kinetic modeling for non-invasive quantification of molecular blood-brain barrier (BBB) transport kinetics.

a Maximum-intensity coronal projections of three 1-s frame dynamic reconstructions (kBq/mL). The extended axial field of view allowed non-invasive measurement of the image-derived input function from the ascending aorta (white outline). b A representative image-derived input function illustrating the importance of high temporal resolution to accurately sample the rapid transport of tracer through the blood pool. c Representative fits to high temporal resolution time-activity curves of ¹⁸F-fluciclovine (FACBC), ¹⁸F-fluorodeoxyglucose (FDG), and ¹¹C-butanol using the adiabatic approximation to the tissue-homogeneity (AATH) model. Fitted curves (teal) were decomposed into their intravascular (red) and extravascular tissue (green) distributions according to the AATH model. d The difference in Akaike Information Criterion (AIC) between the AATH and the standard one-tissue compartment (S1TC) model time-activity curves frame averaged to different intervals (N = 15 samples per tracer from 5 subjects × 3 brain subregions [grey matter, white matter, cerebellum]). Box plot centres indicate average differences and error bars indicate standard deviation of the differences. The AATH model was preferred over the S1TC for 1 to 2 s HTR frame intervals (negative AICs), but not for the 3, 5, 10 s intervals, illustrating the importance of total-body PET in enabling the non-invasive single-tracer BBB PS imaging method. e Representative AATH and S1TC fits to an FDG time activity curve (dashed black line: original) in the grey matter at 1, 5, and 10 s frame intervals, with progressively poorer early peak fitting at greater frame intervals.