Breakthroughs in machine learning have enabled automated recognition of image findings in remarkable ways. This lecture will describe the impact of ML advances on cancer imaging, with a specific focus on the domains where cancer care may be most affected. Strengths and weaknesses of current methods including their scientific bases will be reviewed, as well as practical implications for various real-world clinical practices. The extension of machine learning from full field digital mammography to digital breast tomosynthesis will be specifically explored as an example case, but generalizability to other types of cancer imaging will be described, including implications for reimbursement and patient outcomes.