Deep learning has captured much attention because in many image recognition tasks, it surpasses human performance. The potential for this technology in radiology is substantial, and this talk will describe some of the areas in which deep learning has been or soon will be applied to improve radiological practice.