Effects of dynamic [18F]NaF PET scan duration on kinetic uptake parameters in the knee

Accurately estimating bone perfusion and metabolism using [18F]NaF kinetics from shorter scan times could help address concerns related to patient comfort, motion, and throughput for PET scans. We examined the impact of changing the PET scan duration on the accuracy of [18F]NaF kinetic parameters in the knee.Both knees of twenty participants with and without osteoarthritis were scanned using a hybrid PET-MRI system (53 ± 13 years, BMI 25.9 ± 4.2 kg/m2, 13 female). Seventeen participants were scanned for 54 ± 2 min, and an additional three participants were scanned for 75 min. Patlak Ki and Hawkins kinetic parameters (Ki, K1, extraction fraction) were assessed using 50- or 75-minutes of scan data as well as for scan durations that were retrospectively shortened. The error of the kinetic uptake parameters was calculated in bone regions throughout the knee.The mean error of Patlak Ki, Hawkins Ki, K1, and extraction fraction was less than 10% for scan durations exceeding 30 min and decreased with increasing scan duration.The length of dynamic data acquisition can be reduced to as short as 30 min while retaining accuracy within the limits of reproducibility of Hawkins kinetic uptake parameters.