ABSTRACT
Medical imaging in disease detection, diagnosis, and treatment continues to expand with advances in image quality regulations (e.g., the ACR mammography accreditation program), image acquisition systems (e.g., digital radiography, tomosynthesis, and automated 3D ultrasound), and computerized image analysis (e.g., computer-aided detection [CADe]). e benet of a medical imaging exam depends on both image quality and interpretation quality. Interpreting medical images is the main undertaking of radiologists. However, image interpretation by humans can be limited by incomplete visual search patterns, the potential for fatigue and distractions, the presence of structure noise (camouaging normal anatomical background) in the image, the presentation of subtle and/or complex cancers that require integration of both image data and clinical information, the vast amount of image data in a screening program with low cancer prevalence, and the physical quality of the image itself.
