Short-time total-body dynamic PET imaging performance in quantifying the kinetic metrics of 18F-FDG in healthy volunteers

Purpose To investigate the performance of short-time dynamic imaging in quantifying kinetic metrics of 2-[F-18]-fluoro-2-deoxy-d-glucose (F-18-FDG). Methods Dynamic total-body positron emission tomography (PET) imaging was performed in 11 healthy volunteers for 75 min. The data were divided into 30-, 45- and 75-min groups. Nonlinear regression (NLR) generated constant rates (k(1) to k(3)) and NLR-based Ki in various organs. The Patlak method calculated parametric Ki images to generate Patlak-based Ki values. Paired samples t-test or the Wilcoxon signed-rank test compared the kinetic metrics between the groups, depending on data normality. Deming regression and Bland-Altman analysis assessed the correlation and agreement between NLR- and Patlak-based Ki. A two-sided P < 0.05 was considered statistically significant. Results The 45- and 75-min groups were similar in NLR-based kinetic metrics. The relative difference ranges were as follows: k(1), from 3.4% (P = 0.627) in the spleen to 57.9% (P = 0.130) in the white matter; k(2), from 6.0% (P = 0.904) in the spleen to 60.7% (P = 0.235) in the left ventricle (LV) myocardium; k(3), from 45.6% (P = 0.302) in the LV myocardium to 96.3% (P = 0.478) in the liver; Ki, from 14.0% (P = 0.488) in the liver to 77.8% (P = 0.067) in the kidney. Patlak-based Ki values were also similar between these groups in all organs, except the grey matter (9.6%, P = 0.029) and cerebellum (14.4%, P = 0.002). However, significant differences in kinetic metrics were found between the 30-min and 75-min groups in most organs both in NLR- and Patlak-based analyses. The NLR- and Patlak-based Ki values significantly correlated, with no bias in any of the organs. Conclusion Dynamic imaging using a high-sensitivity total-body PET scanner for a shorter time of 45 min could achieve relevant kinetic metrics of F-18-FDG as done by long-time imaging.