Mean glandular dose estimation using MCNPX for a digital breast tomosynthesis system with tungsten/aluminum and tungsten/ x‐ray anode‐filter combinations

Abstract

Breast cancer screening with x-ray mammography, using one or two projection images of the breast, is an indispensible tool in the early detection of breast cancer in women. Digital breast tomosynthesis (DBT) is a 3D imaging technique that promises higher sensitivity and specificity in breast cancer screening at a similar radiation dose to conventional two-view screening mammography. In DBT a 3D volume is reconstructed with anisotropic voxels from a limited number of x-ray projection images acquired over a limited angle. Although the benefit of early cancer detection through screening mammography outweighs the potential risks associated with radiation, the radiation dosage to women in terms of mean glandular dose (MGD) is carefully monitored. This work studies the MGD arising from a prototype DBT system under various parameters. Two anode/filter combinations (W/Al and W/Al+Ag) were investigated; the tube potential ranges from 20 to 50 kVp; and the breast size varied between 4 and 10 cm chest wall-to-nipple distance and between 3 and 7 cm compressed breast thickness. The dosimetric effect of breast positioning with respect to the imaging detector was also reviewed. It was found that the position of the breast can affect the MGD by as much as 5% to 13% depending on the breast size.

Estimation of Mean Glandular Dose in Mammography Using Voxel Phantom and Monte Carlo Simulation

“…From the results we can see that for both MLO and CC there is an increase in MGD when glandular ratio is higher, which is in agreement with other published data [8]. The results in Figure 5 confirm the linearity of MGD over glandular ratio described in literature [4], [10]. Linearity is valid for all simulated energies as presented in Tables 1 and 2.…”

“…Figures 3 and 4. Resulting values for MGD are lower than those previously reported [8], [9], [10]. The reasons for this are approximations and assumptions used in this work considering modeling of the phantom and source spectrum.…”

Digital Breast Tomosynthesis: State of the Art

“…The variation in the normalized glandular dose coefficient with x-ray tube angle relative to the central projection (0°), referred to as relative glandular dose, was weakly dependent on fibroglandular fraction and x-ray spectrum and showed a larger dependence on breast size and thickness in mediolateral oblique (MLO) view compared with craniocaudal (CC) view (71). Patient positioning, particularly for MLO view, was reported to cause a 5%-13% variation in MGD (72). While the initial study (71) was based on a DBT system using molybdenum or rhodium target xray tube, subsequent studies addressed tungsten target x-ray tubes (72,73).…”

“…Patient positioning, particularly for MLO view, was reported to cause a 5%-13% variation in MGD (72). While the initial study (71) was based on a DBT system using molybdenum or rhodium target xray tube, subsequent studies addressed tungsten target x-ray tubes (72,73). Currently, many DBT systems use tungsten target with aluminum, silver, or rhodium filter (Table 1) and operate at a slightly higher applied tube voltage (kilovolts peak) than FFDM, partly to reduce radiation dose to the breast.…”

Objective models of compressed breast shapes undergoing mammography

“…In a few studies, models of the breast compressed for the medio-lateral oblique (MLO) view have been used, but the models were subjectively derived. 18,19,[29][30][31] Although these models have proved appropriate, the development of more sophisticated models that better represent not only the "average" breast shape but also attempt to include the span of clinically realistic breast shape variations would be useful in many of the areas of research listed above that use software breast models.…”