Measurement of the linear attenuation coefficient of breast tissues using polienergetic x-ray for energies from 12 to 50 keV and a silicon dispersive detector

Abstract

In this work we obtained experimental linear attenuation coefficients, at mammography energy range (12–50 keV), for 57 breast tissue samples, previously classified as adipose tissue (17), glandular tissue (9) or invasive ductal carcinoma (IDC, grades 1 (3), 2 (12) or 3 (16)). The density of those samples was measured using the Archimedes Principle. A polienergetic narrow beam was obtained using a tungsten anode x-ray tube, collimators (before and after the sample) and silicon drift detector. The linear attenuation coefficient of known composition materials was obtained to validate the experimental setup, showing differences up to 3%, lower than the uncertainties of the measurements (4%). The density of the samples was compared with previous published data. Our results show that the density of IDC tissue increases for higher grades. The variability of the linear attenuation coefficient was also studied, for different regions of the same sample (4%) and for each sample group (18–25%). The elemental composition of the tissues is the main responsible for the group variability. Analyzing the mean results for the different groups, the adipose tissue presents the lowest linear attenuation coefficient, then the glandular tissue, the IDC 1, IDC 2 and IDC 3 being the largest. A statistical analysis was performed on the linear attenuation coefficient of the tissue groups. Adipose tissue group was statistically different from the other groups of tissues, for all the energy interval. There was no statistical difference of the glandular tissue group from any of the IDC groups, for all the energy interval, due to the variability in density and elemental composition. The group of IDC grade 2 was not statistically different from the IDC grade 3 group, however there is a tendency that the linear attenuation coefficient of IDC grade 3 is slightly higher than the IDC grade 2 one.