학회지검색
대한유방검진의학회 학회지
Review Article
1 Department of Radiology, Severance Hospital, Research Institute of Radiological Science Yonsei University, College of Medicine, 2Department of Radiology, Mokdong Hospital, Ewha Womans University, College of Medicine, 3 Department of Radiology, Kyungpook National University Chilgok Hospital, School of Medicine, Kyungpook National University, 4BeamWorks Inc., 5School of Computer Science and Engineering, Kyungpook National University
Artificial intelligence (AI) systems for breast US are now commercially available for clinical use. AI for breast US has high performances, but both breast US and AI systems are used in a variety of clinical settings that can affect outcomes of using AI for decision support. Among the studies using commercially available AI systems for breast US, studies commonly show that the diagnostic performances are significantly improved by using AI. There are several factors that users should consider for appropriate AI use such as the use case of the AI system, how the AI system generates results, and the differences in the level of suspicion given by different AI systems. For optimal AI integration, differences in workflow (technician vs. radiologist performed examinations) and clinical background (screening vs. diagnostic setting) should be considered. In this review, we will discuss the status of AI systems for breast US, the outcomes of using AI for breast US interpretation and what we need to consider when we use AI in our practice.
Key Words: Breast; Ultrasonography; Artificial intelligence