Characterization of the scatter component in large axial field-of-view PET scanners: a Monte Carlo simulation study

A simulation study was conducted to estimate the scatter fraction (SF) and to determine the most suitable line source radial offset displacement required to measure the SF for the total body long axial field-of-view (LAFOV) PET scanner by simulating different cylindrical and anthropomorphic digital phantoms. Simulations are conducted using a scanner model together with scatter phantoms adapted from the NEMA NU-2 2007 scatter phantom design, modified to suit the dimensions of the respective scanners, with line sources of various activities placed at a 45 mm radial offset. SF estimates obtained using the NEMA protocol are compared to values obtained with uniformly filled water phantom of the same length. A whole-body study is conducted using a set of 12 anthropomorphic phantoms with different BMIs, with different organs and anatomical structures filled with realistic concentrations of
¹⁸
F-FDG. The SF obtained at 45 mm radial offset using 1 kBq/ml with the 200 cm (LAFOV scanner) and 70 cm (mCT) cylindrical phantoms are 40.07% and 34.35%, respectively. In both cases, a comparison with the SF estimate obtained with a uniformly filled cylindrical phantom shows that the NEMA NU2-2007 phantom with the line source positioned at the recommended radial offset of 45 mm significantly overestimates the SF. Instead, it was found that for both scanners, the optimal radial offset for accurate estimation of the SF was approximately 60 mm. High SF correlation coefficients were obtained between the SFs estimated with anthropomorphic phantoms with realistic biodistribution of
¹⁸
F-FDG and an equivalent volume cylindrical phantom for the LAFOV PET scanner; in addition, BMI was strongly positively correlated with SF. The SF is found to be higher for the LAFOV compared with the mCT PET scanner. The optimal radial displacement for a LAFOV PET scanner using a NEMA-like phantom was found to be 60 mm, compared to the value of 45 mm suggested by the NEMA protocol.