A survey of approaches in Deep Learning techniques for the detection and classification of mammography abnormalities

Mammography is currently the most widely used laboratory study for the early detection of precursor abnormalities of breast cancer, which is one of the main causes of mortality worldwide since this disease generates the appearance of large volumes of cancerous cells in the breast.Thus, Computer-aided detection or diagnosis (CADe and CADx) can help the specialist improve the accuracy of the diagnosis, providing the patient with a better treatment and disease management assessment. Unfortunately, there are several circumstances in which it is possible to obtain an erroneous diagnosis, such as subjectivity on the part of the radiologist due to the size or morphology of the anomalies detected, as well as the fact that mammograms may include embedded noise.For this reason, this manuscript seeks to provide the reader with a more general overview of how this issue has been addressed in recent years (2016 onwards) through the compilation of various scientific research focused on the detection and classification of breast cancer precursor anomalies based on the implementation of artificial intelligence techniques, where it is possible to visualize the performance of each of the proposed models through the implementation of various metrics such as F1 Score, AUC, accuracy, etc. Thus, these studies have been obtained from recognized multidisciplinary scientific sites in the world such as Nature, Springer, IEEE Xplorer, and PubMed, among others.