Abstract
Since 2017, we have been conducting breast cancer screening combined with tomosynthesis imaging as part of opportunistic screening.
The addition of digital breast tomosynthesis (DBT) improved the close recall rate, cancer detection rate, and positive predictive value of breast cancer screening, but it posed the problems of increased exposure dose and imaging time. Therefore, we examined whether it is possible to interpret mammography using only synthesized 2D images (S2D) from DBT imaging (3D) data, and whether it is possible to omit 2D mammography imaging (2D).
We used the images of 2,611 women who had undergone mammography screening with the addition of DBT, and the images of 5,211 breasts, including the breasts of only one side in some cases. These images were set to the monitor so that S2D + 3D (Group A), 2D + 3D (Group B), and 2D only (Group C) images were displayed, and six doctors with screening mammography interpretation certification A or AS interpreted the images, with staggered timings. Image groups A, B, and C were interpreted separately, and it was assessed whether there were some differences in the process indication among the three groups. The results revealed no differences in the categorical judgment among the three groups. In regard to the process indicators, in groups A, B, and C, the recall rates were 7.1%, 7.8%, 9.1%, the cancer detection rates were 0.27%, 0.23%, 0.19%, and the positive predictive values were 3.8%, 2.9%, 2.1%, respectively. In conclusion, process indication using synthesized 2D imaging was better than the interpretation of 2D images alone. For the future, it is suggested that omission of 2D imaging may be useful in opportunistic breast cancer screening conducted with the addition of DBT.
