Segmentation-based MR attenuation correction including bones also affects quantitation in brain study: An initial result of [18F]FP-CIT PET/MR for patients with parkinsonism

Objectives The aim of this study is to evaluate brain PET/MR on quantification of [18F]FP-CIT binding, focusing on effects of ultrashort echo time (UTE)-based AC including bone segmentation.

Methods Ten patients with initially suspected of Parkinsonism were prospectively enrolled. Emission scan was acquired 110 min after FP-CIT injection on dedicated PET/MR, immediately followed by another emission scan on PET/CT. UTE-based attenuation map was used to classify the voxels into 3 tissues, bone, soft tissue and air. Specific-to-nonspecific dopamine transporter (DAT) binding ratio (BR) was calculated in the striatum using statistical probabilistic anatomical mapping. The level of agreement was assessed with intraclass correlation coefficients (ICC). Voxelwise comparison between PET/MR and PET/CT was performed. We compared non-AC images to analyze UTE-based AC effects on DAT quantification.

Results BR in the putamen obtained from PET/MR and PET/CT showed low inter-equipment variability, while BR in the caudate nucleus showed significant variability (ICC = 0.955 and 0.665, respectively). BR in the caudate nucleus was significantly underestimated by PET/MR compared to PET/CT (p < .0001). Voxelwise analysis revealed PET/MR showed significantly low BR in the periventricular regions, which was caused by misclassification of ventricle as air on attenuation map. We also compared non-AC images, which revealed low inter-equipment variability even in the caudate nucleus (ICC = 0.951 and 0.904, respectively).

Conclusions Our data demonstrates spatial bias of the DAT BR in FP-CIT PET/MR. Voxelwise analysis and comparison of non-AC images implied misclassification of ventricle as air account for the bias. To obtain reliable quantification for brain PET/MR studies, alternative segmentation strategies are warranted.