Key points are not available for this paper at this time.
To lower breast cancer morbidity and mortality, millions of breast imaging tests are carried out year. Breast imaging tests are carried out for cancer screening, diagnostic evaluation of suspicious findings, assessing the severity of the illness in patients who have just been diagnosed with breast cancer, and assessing treatment response. However, the interpretation of breast imaging can be arbitrary, laborious, slow, and open to human error. Deep learning (DL) has a great potential to perform medical imaging tasks at or above human-level performance and may be used to automate parts of the breast cancer screening process, increase cancer detection rates, reduce needless call-backs and biopsies, improve patient risk assessment, and create new opportunities for disease prognostication. Retrospective and small reader studies support this claim. In order to verify these suggested tools and open the door to actual therapeutic application, prospective studies are urgently required. To meet the distinct ethical, medico-legal, and quality control challenges that DL algorithms provide, new regulatory frameworks must also be created. In this paper, we cover the fundamentals of DL, present current DL breast imaging applications, such as cancer diagnosis and risk prediction, and talk about the difficulties and potential paths for AI-based breast cancer systems.
Ghosh et al. (Fri,) studied this question.