X-ray tomosynthesis techniques for breast imaging

In this work, we present a systematic approach to improving image quality of x-ray digital breast tomosynthesis (DBT). DBT is increasingly used for breast cancer screening and diagnosis with its advantages over mammography being clinically evidenced. Image reconstruction algorithms play an important role in providing quality images for such diagnosis. It is desirable for the algorithms to enhance image contrast and resolution and to suppress noise and image artifacts. Using from one-shot analytic image reconstruction algorithms and iterative algorithms to deep-learning-enabled approaches, we systematically provide the roles of such algorithms in improving image quality.