Application of an anatomical brain phantom in the evaluation of partial volume effect correction of amyloid PET studies

to simulate amyloid PET (A-PET) studies using an anatomical brain phantom (StepBrain—SB) with varying gray-to-white matter ratios (GWR), to quantify partial volume effects (PVE)-related bias and assess the accuracy of PVE correction (PVEc) methods. SB was scanned to simulate clinical A-PET studies across GWR from 0 to 120%. GWR and Standardized Uptake Value Ratios (SUVR, normalized by cerebellum) derived from an atlas-based volume-of-interest set were analyzed before and after PVEc using 2-compartment (2C), 3-compartment (3C), modified 3C (m3C), and geometric transfer matrix (GTM) methods. Accuracy and precision of PVEc methods were compared by linear mixed models analysis. Simulated data reproduced the clinical range of amyloid uptake patterns. Accuracy and precision differed significantly among PVEc methods for both GWR and SUVR (F>95 for all). Uncorrected GWR and SUVR showed 14.1% and 19.8% mean absolute inaccuracy, respectively, which decreased up to 6.1% and 7.6% when applying the PVE correction methods, with m3C and GTM providing the highest accuracy and precision. The SB phantom reproduced A-PET distributions and provided a ground truth for validation of PVEc methods. 3C-, m3C-, and GTM-based PVEc significantly improved tracer ratio accuracy without relevant precision loss, supporting their use for quantitative analysis of clinical A-PET studies.