Mammographic imaging performance: Histogram quantitative metrics using Monte Carlo generated images

One of the major limitations of mammography is the detectability of low contrast inhomogeneities in breast parenchyma. The influence of the exposure parameters on the detectability of lesion inhomogeneities is investigated by means of Monte Carlo simulation studies. Breast was simulated utilizing a mathematical semicylindrical phantom with physical properties corresponding to glandular breast tissue. Specifically, the radius and thickness of the breast phantom were set to 4 cm. Microcalcifications were mathematically simulated utilizing a rectangular parallelepiped of projective dimensions of 0.5 cm times 0.5 cm, with varying thickness and composition corresponding to calcium oxalate (CO). The phantom was virtually irradiated with a range of commonly used mammographic spectra, using Monte Carlo simulation to describe the photon transport phenomena. The produced image quality, described based on characteristics derived from the Gaussian distribution of the greylevels of the signal, strongly depends on the anode and filter material. The proper selection of the exposure parameters could further improve the diagnostic capability of mammography.