Design of a 3D Mammography System in the Age of Personalized Medicine
Abstract
Despite the latest move toward 3D breast cancer screening with digital breast tomosynthesis (DBT), current systems continue to suffer from blurring in calcification imaging. We performed virtual clinical trials (VCTs) of anthropomorphic phantoms to analyze calcification detection in DBT. Additionally, we investigated super-resolution (SR) as a mechanism to improve calcification visibility. It is known from previous work that there are anisotropies in SR in current systems and that these anisotropies can be eliminated with the use of detector motion. We have recently shown that it is possible to eliminate the anisotropies by re-arranging the source positions in clusters with submillimeter spacing. This new design eliminates the need for detector motion and hence the need for a thick detector housing, which is cumbersome for patients. We have also modeled SR in magnification imaging, and showed that SR is optimized by circular source motions as opposed to linear motions. In addition, we analyzed how personalized acquisition geometries, customized around the size and shape of the breast, improve dense tissue visualization. While our early work was focused on Defrise phantoms, this analysis has recently expanded to anthropomorphic phantoms. We validated this phantom model as having realistic textural detail compared against clinical data.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 2020
- Accession Number
- AD1104162
Entities
People
- Raymond J Acciavatti
Organizations
- University of Pennsylvania