Fig. 2

Differential activity concentration volume histograms (dAVH). ⁹⁰Y–PET images were reconstructed using a GE Discovery ST PET/CT system with 2.73 × 2.73 × 3.27mm³ detector elements. A full 3D-OSEM VUEPoint HD reconstruction algorithm, two iterations, 15 substeps, was used. Acquisitions at day 1 and 4 were performed using 15 min scan time, while 30 min acquisition time was used for imaging at days 5, 6, and 12. Acquisition duration was varied over time as we aimed to perform acquisition in realistic clinical conditions. Scan time below 30 min are very unlikely in the clinical practice for activity concentrations below 2 MBq/mL, especially in the absence of a TOF scanner. Activity distributions of ⁹⁰Y–PET images were analyzed in terms of dAVHs. Positive skewness (indicated with s) was obtained for all dAVHs. a–e dAVH of the cylindrical insert fixed into the liver cavity of the anthropomorphic phantom. Heterogeneous activity distributions were obtained for all acquisitions. f–j dAVH of the liver background. Visual inspection of the images (k–o) showed the presence of hot and cold spots in the insert, supposed to be uniformly filled with ⁹⁰Y. As a general rule, both for the tumor insert and the liver compartment the lower the activity concentration, the higher the skewness value