Experimental and simulation studies for the optimization of dedicated scintimammography cameras

Abstract

The main drawback of scintimammography with conventional gamma cameras is the large distance between detector and breast, which limits imaging performance and adds significant background noise from the heart and other organs. These limitations can be minimized by dedicated gamma cameras placed in tight contact with the breast. The purpose of this study was to determine the lowest Tumor-to-Background (T:B) ratio for detectable lesions and the smallest lesions that can be imaged with a dedicated system. The system is based on an R3292 position sensitive photomultiplier tube (PSPMT) from Hamamatsu, coupled to a 10 cm in diameter CsI(Tl) crystal and a general purpose collimator with hexagonal holes. Two crystals were assessed, with pixel size of 2 x 2 mm(2) and 3 x 3 mm(2). The experimental setup consists of a cylindrical breast phantom, cylindrical tumor phantoms positioned inside the breast phantom at various depths, and a heart phantom. The same setup, as well as some additional tumor phantoms, were modeled with the GATE simulation toolkit, following the validation of the entire system. A 20% energy window and the Dual Energy Window Subtraction Method (DEWSM) for scatter correction were applied in experimental and simulation data. In all measurements Tc-99m was used. Signal to Noise Ratio (SNR) and contrast were measured in all experiments and used for quantification of results. Experimental and simulation results show that the lowest T:B ratio that can be detected is 2:1 with tumor phantom volume >= 0:25 ml at 1 cm depth. The smallest volume of a tumor phantom that can be detected is 0.04 ml with T:B ratio equal to 9:1 at 1 cm depth.