Room: AAPM ePoster Library
Purpose: The anti-scatter grid performance was studied on three 2D mammography units for a range of breast phantom thicknesses and commonly used techniques.
Methods: The mammography units and associated grids studied were Senographé Pristina and Essential (GE Healthcare, Waukesha, WI) and Selenia Dimensions (Hologic, Marlborough, MA). The ratios of the units’ grids were 11:1 for the Pristina, 5:1 for the Essential and 6:1 for the Dimensions. Two, 4, 6 and 8 cm thicknesses of CIRS (Computerized Imaging Reference Systems, Norfolk, VA) Model 014A 50% glandular/50% adipose phantom were employed. Scatter measurements were made with the phantom on the grid cover. Primary measurements were made with the phantom positioned directly beneath the x-ray tube and collimation assembly, and pre and post phantom collimation were employed to achieve a 1 cm diameter beam. Values from the raw flat field and gain corrected images were used to calculate the following: scatter-to-primary ratios with and without the grid; grid primary and scatter transmission, and grid contrast-to-noise improvement factor for the grid of each unit and for each phantom thickness and technique.
Results: Relative intensity of scatter imaged with and without grid increased with increasing thickness. At 4 cm all three units and grids had contrast-to-noise improvement factors (CNIFs) of ˜ 1.0 and comparable performance. The CNIF on all three systems was slightly less than 1.0 for thicknesses < 4 cm and was greater than 1.0 for > 4 cm with significant CNIF Improvement at 8 cm.
Conclusion: Hologic Unit and 6:1 grid performed better at 2 cm thickness while the GE Pristina Unit and 11:1 grid performed better at 8 cm. Our results also imply that employing a grid in Digital Breast Tomosynthesis would improve image quality for the majority of women.
Not Applicable / None Entered.