Purpose: To investigate how the displayed average glandular dose (AGD) of Hologic DBT systems compares to computed AGD, and how this comparison is modulated by thickness, and the assumptions in women breast glandularity.
Methods: For 2, 4, 6, and 8 cm of BR-12, exposures were acquired using the Auto-Filter AEC mode on six Hologic DBT systems. Entrance exposures and HVL were measured using a RaySafe X2 solid state detector, and the displayed entrance skin dose (ESD), displayed AGD, and displayed exposure index (EI) were recorded. The AGD was first estimated under the assumption of 50% glandularity, 50% adipose tissue composition. It was then computed using the assumption of breast composition for the average woman aged 40-49 years old (100%, 65%, 35% and 14% glandularity for 2-, 4-, 6- and 8-cm breast thickness, respectively). The AGD computed using these assumptions was compared to the displayed AGD.
Results: Under the assumption of 50% glandularity, the absolute percentage difference between the computed AGD and displayed AGD ranged from 0.3% (for 2 cm of BR-12 in 3D) to 20.5% (2 cm in 2D). The deviation is most apparent at the largest thickness in 3D, where the displayed AGD overestimates the computed AGD.Using the assumption of glandularity based on the average woman aged 40-49 years old, the absolute percentage difference between the computed AGD and displayed AGD ranged from 0.04% (2 cm in 2D) to 23.2% (8 cm in 2D). The deviation is most apparent at the largest thickness in both 2D and 3D mode, where the displayed AGD underestimates the computed AGD.
Conclusion: This study suggests that the discrepancy between displayed AGD and computed AGD may be correlated with the breast thickness. More data is needed to support this conclusion. The displayed AGD may overestimate or underestimate AGD based on the breast composition.
Not Applicable / None Entered.