Deep learning in breast radiology: current progress and future directions

摘要: This review provides an overview of current applications deep learning methods within breast radiology. The diagnostic capabilities in radiology continue to improve, giving rise the prospect that these may be integrated not only into detection and classification lesions, but also areas such as risk estimation prediction tumor responses therapy. Remaining challenges include limited availability high-quality data with expert annotations ground truth determinations, need for further validation initial results, unresolved medicolegal considerations. KEY POINTS: • Deep (DL) continues push boundaries what can accomplished by artificial intelligence (AI) imaging distinct advantages over conventional computer-aided detection. DL-based AI has potential augment radiologists improving accuracy, increasing efficiency, supporting clinical decision-making through prognosis therapeutic response. DL implementation a paucity prospective on utilization yet questions regarding utilization.