Monte Carlo simulation studies for the determination of microcalcification thickness and glandular ratio through dual-energy mammography

Abstract

The majority of breast carcinomas can be associated to the presence of calcifications before the development of a mass. However, the overlapping tissues can obscure the visualization of microcalcification clusters due to the reduced contrast-noise ratio (CNR). In order to overcome this complication, one potential solution is the use of the dual-energy (DE) technique, in which two different images are acquired at low (LE) and high (HE) energies or kVp to highlight specific lesions or cancel out tissue background. In this work, the DE features were computationally studied considering simulated acquisitions from a modified PENELOPE Monte Carlo code. The employed irradiation geometry considered typical distances used in digital mammography, a CsI detection system and an updated breast model composed of skin, microcalcifications and glandular and adipose tissues. The breast thickness ranged from 2 to 6cm with glandularities of 25%, 50% and 75%, where microcalcifications with dimensions from 100 up to 600μm were positioned. In general, results pointed an efficiency index better than 87% for the microcalcification thicknesses and better than 95% for the glandular ratio. The simulations evaluated in this work can be used to optimize the elements from the DE imaging chain, in order to become a complementary tool for the conventional single-exposure images, especially for the visualization and estimation of calcification thicknesses and glandular ratios.