Computational evaluation of the bucky components influence on the estimation of normalized glandular dose in digital mammography

Author(s)

Godeli, J. ; Cunha, D. M. ; Potiens, M. P. A. ; Poletti, M. E.

• Document type: Journal article
• Source: Radiation Physics and Chemistry; v. 222, p. 8-pg., 2024-06-12.

Abstract

The mean glandular dose (MGD) is the most suitable dosimetric quantity used in mammography to describe the absorbed dose by the breast, although it cannot be directly acquired. Studies have provided conversion factors widely implemented in international dosimetry protocols to estimate MGD, such as normalized glandular dose (DgN). Over time, the DgN estimation was refined by considering geometric models that approach a real clinical environment, such as new anode/filter combinations, compression plate and breast models. However, there is no detailed study of how the bucky (support plate, antiscatter grid and detector) can affect the DgN estimation. A modified PENELOPE Monte Carlo code was used for DgN estimation. The irradiation geometric model was built as a complete digital mammography system, considering a homogeneous breast and different typical bucky models in commercial mammography units. Simulations were carried out for mono and polyenergetic beams considering different imaging geometries. Studies with monoenergetic beams showed that the bucky presence affected DgN mainly for higher beam energies and thinner breasts. The breast support plate was the bucky component that most affected the DgN, followed by the antis-scatter grid and finally, the image detector. Studies with polyenergetic conventional (low-energy) spectra showed that the bucky exerted a minimal influence on DgN values (less than 1.0%). For high-energy spectra, mainly employed in modalities such as contrast-enhanced digital mammography, the DgN values were more affected by the bucky, increasing by 4.8% the DgN values for a 2 cm thick breast and a W/Cu 50 kV spectrum. Bucky inclusion in computer simulations is highly recommended mainly for thinner breasts and high-energy spectra. To simplify the simulations, we confirm that a homogeneous carbon fibre block support, with thickness between 3.9 and 4.1 mm, can be used as a good substitute for a complete bucky model. (AU)

• FAPESP's process: 18/05982-0 - R&D&I in radiation metrology in the health area
• Grantee: Linda Viola Ehlin Caldas
• Support Opportunities: Research Projects - Thematic Grants