Trimodel Mammography with Perfect Coregistration

This project aims at developing a trimodel x-ray mammography imaging system to improve both sensitivity and specificity in breast cancer screening and diagnosis, particularly for radiologically dense breasts. In the proposed system, three complementary image datasets will be generated from a single data acquisition the first is the conventional absorption contrast mammography image, the second is a novel phase contrast mammography image with enhanced edges and reduced anatomical background, the major confounding factor in reading mammography the imaging characteristics suggest that this contrast mechanism would be preferable for cancer mass detection. The third image is the dark-field mammogram, which is sensitive to the local distribution of microcalcifications, calcified vessels, and other small objects in the breast. The proposed system will be constructed, optimized, and evaluated using mastectomy specimens.