Impact of target-to-background ratio, target size, emission scan duration, and activity on physical figures of merit for a 3D LSO-based whole body PET/CT scanner.

The aim of our work is to describe the way in which physical figures of merit such as contrast-to-noise ratio (CNR) behave when varying acquisition parameters such as emission scan duration (ESD) or activity at the start of acquisition (A(acq)) that in clinical practice can be selected by the user, or object properties such as target dimensions or target-to-background (T/B) ratio, which depend uniquely on the intrinsic characteristics of the object being imaged. Figures of merit, used to characterize image quality and quantitative accuracy for a 3D-LSO based PET/CT scanner, were studied as a function of ESD and A(acq) for different target sizes and T/B ratios using a multivariate approach in a wide range of conditions approaching the ones that can be encountered in clinical practice. An annular ring of water bags of 3 cm thickness was fitted over an IEC phantom in order to obtain counting rates similar to those found in average patients. The average scatter fraction (SF) of the modified IEC phantom was similar to the mean SF measured on patients with a similar scanner. A supplemental set of micro-hollow spheres was positioned inside the phantom. The NEMA NU 2-2001 scatter phantom was positioned at the end of the IEC phantom to approximate the clinical situation of having activity that extends beyond the scanner. The phantoms were filled with a solution of water and 18F (12 kBq/mL) and the spheres with various T/B ratios of 22.5, 10.3, and 3.6. Sequential imaging was performed to acquire PET images with varying background activity concentrations of about 12, 9, 6.4, 5.3, and 3.1 kBq/mL, positioned on the linear portion of the phantom's NECR curve, well below peak NECR of 61.2 kcps that is reached at 31.8 kBq/mL. The ESD was set to 1, 2, 3, and 4 min/bed. With T/B ratios of 3.6, 10.3, and 22.5, the 13.0, 8.1, and 6.5 mm spheres were detectable for the whole ranges of background activity concentration and ESD, respectively. The ESD resulted as the most significant predictor of CNR variance, followed by T/B ratio and the cross sectional area of the given sphere. Only last comes A(acq) with a weight more than halved with respect to ESD. Thus, raising ESD seems to be much more effective than raising A(acq) in order to obtain higher CNR, which is the physical figure of merit closely related with target detectability, at least in the simple task of the signal known exactly background known exactly model.