Constant-Activity-Rate Infusions for Myocardial Blood Flow Quantification with 82Rb and 3D PET

Quantification of myocardial blood flow (MBF) can be performed using dynamic ⁸²Rb PET imaging. However, the small dynamic range of some 3D PET systems can prevent accurate measurement of the first-pass activity with standard ⁸²Rb bolus infusions. The purpose of this study was to investigate the use of slow constant-activity-rate infusions to optimize MBF quantification with 3D PET. Methods: Dynamic 3D PET imaging was performed in a normal dog with 150 MBq of ⁸²Rb infused over 15, 30, 60, 120 and 240 s. Left ventricular mean MBF was quantified using a one-compartment model of ⁸²Rb kinetics, and compared to standard values obtained using ¹³N-ammonia with a two-compartment model. Results: Peak deadtime and coincidence count-rates decreased by almost 50% and 70% respectively with 240 vs. 30 s tracer infusion, but integral true coincidences (prompt - delayed) were maintained. Mean MBF values remained accurate (1.4 vs. 1.3 mL/min/g) using 240 vs. 30 s infusions, and vs. standard ¹³N-ammonia MBF values (1.3 mL/min/g). Precision of the MBF estimates was also improved by a factor of 3.5. Conclusion: Constant-activity-rate infusions reduce the dynamic range needed for MBF quantification with ⁸²Rb and 3D PET imaging. Using a one-compartment model, the precision of MBF estimates are also improved with slower infusions. This may permit improved MBF quantification using 3D PET, with high integral counts maintained for conventional perfusion imaging.