RT Journal Article SR Electronic T1 Optimized 82Rb infusion profile for flow quantification with low count-rate PET system using a physical flow phantom JF Journal of Nuclear Medicine JO J Nucl Med FD Society of Nuclear Medicine SP 1944 OP 1944 VO 57 IS supplement 2 A1 Hanif Gabrani-Juma A1 Owen Clarkin A1 Robert deKemp A1 Ran Klein YR 2016 UL http://jnm.snmjournals.org/content/57/supplement_2/1944.abstract AB 1944Objectives Dynamic imaging of rubidium-82 PET is cost-effective and convenient for routine clinical myocardial blood flow quantification. However, due to its short radionuclide half-life (76 sec), a wide dynamic range of activities (>4 orders of magnitude) are present in the scanner field of view during the time-course of imaging. PET scanners can be saturated by early bolus activity or have insufficient count statistics for late frame image reconstruction; this is especially problematic for low count-rate systems. We hypothesize that administration of 82Rb using a constant activity-rate ‘slow-bolus’ infusion can achieve accurate flow quantification on a low count-rate based PET system.Methods A water perfused compartment exchange phantom was imaged using 82Rb on a BGO PET/CT scanner (GE D600) with injected activities varying from 625 - 2400 MBq and constant activity-rate infusion times between 30 - 120 seconds. Peak dead-time correction factor (DTF) for each series was recorded as a potential indicator for camera saturation. Regions of interest were defined on the inlet and exchange chambers to sample input and output time-activity-curves (TACs) respectively. A 1-tissue-compartment model with wash in and wash out parameters (K1 and k2) was fitted to the TACs. Wash in and wash out flow rates were compared to physical flow meter values (truth) to determine quantification biases, with a <5% bias assumed acceptable.Results There was good agreement between the activity rate (MBq/s) versus peak DTF (r2 = 0.82). K1 was not affected by changes in activity rate. No bias in k2 was present at low activity rates, but significant bias was noted for activity rates corresponding to peak DTF greater than 2.2 (55% dead-time). Maximum activities with <5% bias were estimated at 975, 1525 and 2400 MBq for 30, 60 and 120-second infusions respectively.Conclusions The amount of activity used for dynamic 82Rb imaging was successfully increased using ‘slow-bolus’ infusions, enabling accurate flow quantification and higher count statistics for late time-frame imaging. Peak DTF values can be used for quality control of camera saturation and blood flow quantification bias.