Digitalization of a Mammographic Phantom View through a Monte Carlo Simulation

The aim of this study is the development of a methodology to reconstruct the image of detail defects from a mammographic phantom employed in quality control testing. The MCNP-4c2 code has been used to model the mammography unit and the CIRS 11A (MAMMO PHANTOM SP01) mammographic phantom. This phantom is made of poly-methyl methacrilate and contains a reference point, contrast and resolution detail targets, groups of microcalcifications and fibres. The image in a mammogram is produced by the unscattered fraction of the emitted X-rays, and the scattered fraction, not useful to imaging purpose. The unscattered fraction has been scored with a photon flux tally (F5), employing a point detector matrix below the mammographic phantom. The MCNP code has been modified and recompiled to let a large number of tallies per input file. Point detectors have an inherent singularity with the square distance from the point of interest to the detector. Therefore, the use of a point detector tally to score both fluxes has a well behaviour only with the uncollided fraction (variance zero). Instead of scoring with a point detector the scattered fraction, it has been used a track length estimate tally (F4), modeling a lattice of rectangular cells. Photons have been forced to reach each one of the lattice cells with a matrix of DXTRAN spheres, centred upon them. The characteristic curve of the mammographic film has been obtained throughout several experimental measures of grey levels with different aluminium layers and Monte Carlo simulations, obtaining a response function between output air kerma and grey level. The simulated images agree with the range of values, indicating that this method will be suitable for training purposes or phantom designing.