Patient-specific Breast Dosimetry in Breast Cancer Screening

Although mammography has been performed on millions of women a year for decades, we still do not have an accurate method to estimate the radiation dose each specific patient receives during her mammogram. The dose metric used in mammography reflects the dose received by a breast approximated to consist of a homogeneous mixture of adipose and glandular tissue, with a standard shape, and compressed for the cranio-caudal view only. We have recently shown that the error introduced by the homogeneous mixture assumption alone can be as high as 100%. With the introduction of screening tomosynthesis, for the first time we will have available a pseudo-three dimensional representation of each patient’s true glandular tissue structure and distribution. Therefore, the objective of this research project is to develop and test the methods necessary to obtain an accurate estimate of the personalized radiation dose each patient receives during breast cancer screening with tomosynthesis imaging, and with mammography if performed in conjunction with tomosynthesis. This will result in the ability to obtain a personalized dose estimate for each of the millions of breast cancer screening exams performed each year on women around the world.

Funding Agency: Susan G. Komen Foundation for the Cure

Grant Type: Challenge Grants: Investigator-Initiated Research

Funding Period: 08/16/13-08/15/17

Figure 2

Initial results of automated tissue classification of breast tomosynthesis images – the first step in achieving personalized radiation dose estimates in breast cancer screening.