Comparison between average glandular dose (AGD) calculated by mammography equipment and VolparaDose software

In mammography equipment, the average glandular dose (AGD) is calculated from the incident air kerma (Ki) multiplied by conversion coefficients derived from Monte Carlo simulations, which consider breast thickness and density. However, calculating AGD using specific and true patient information results in a dose that is closer to the real dose. This study compares the AGD calculated by two different methods: the equipment, which uses conversion coefficients, and the VolparaDose software, which uses the patient-specific volumetric breast density (VBD). The study was conducted with a sample of 3,209 images from screening mammography exams. Through hypothesis tests for the mean difference, it was demonstrated that for compressed breast thicknesses > 27.1 mm, the AGD was significantly lower (p-value < 0.05) when calculated by the equipment itself compared to the AGD calculated by the VolparaDose software. The VBD was a significant factor in the difference in AGD calculated between the two methods. The results suggest that the AGD calculated by the equipment may be underestimating the dose when compared to the AGD calculated by the VolparaDose software.