Quantification of Low-Dose Dynamic [18F]FDG Brain PET on a Long Axial Field-of-View Scanner

Purpose Long axial field-of-view (LAFOV) PET scanners offer increased sensitivity, enabling reduced injected activity. While dose reduction has been explored for whole-body imaging using SUV metrics, its applicability to dynamic brain PET kinetic analysis remains unclear. This study evaluated whether injected activity can be reduced in dynamic brain [ 18 F]FDG LAFOV PET while preserving the quantitative accuracy of kinetic parameters. Methods Nine subjects underwent 65-minute dynamic LAFOV PET following injection of 3 MBq/kg [ 18 F]FDG. Dose-reduced datasets (1.5, 0.75, and 0.375 MBq/kg) were simulated via list-mode undersampling. Two-tissue compartment modelling and Patlak analysis were performed across multiple brain regions. Kinetic parameters from reduced-count data were compared with full-count data using relative percentage error. Results K i estimates were robust to dose reduction. Errors remained below 5% at 1.5 MBq/kg and below 10% at 0.75 and 0.375 MBq/kg for both modelling approaches, with only slightly higher percentual errors for smaller regions (< 11 mL) at the lowest injected activity. In contrast, the microparameters ( K 1 , k 2 , k 3 ) and V B exhibited larger, region size-dependent biases. For large regions (> 35mL), percentual errors generally remained within 10% down to 0.375 MBq/kg. For medium-to-small-sized regions (< 35 mL) percentual errors increased substantially, remaining within 10% only at 1.5 MBq/kg and rising to 15-25% at 0.75 MBq/kg and 20–40% at 0.375 MBq/kg. Conclusion Reliable estimation of K i is feasible even at the lowest undersampled activity level. However, accurate quantification of microparameters is limited at low count levels: for small regions, dose reduction below 3 MBq/kg is not recommended, whereas for larger regions reductions to 1.5 MBq/kg may be acceptable. Sensitivity and parameter accuracy could likely be improved by repositioning the head closer to the scanner’s axial centre. These findings provide practical guidance for designing low-dose dynamic [ 18 F]FDG brain PET protocols on LAFOV systems.