State-of-the-Art in Responsible, Explainable, and Fair AI for Medical Image Analysis
Soheyla Amirian, Fengyi Gao, Nickolas Littlefield, Jonathan Hill, Adolph J. Yates, Johannes F. Plate, Liron Pantanowitz, Hooman H. Rashidi, Ahmad P. Tafti
Abstract
Integrating responsible, explainable, and fair artificial intelligence (REF-AI) into medical image analysis has gained significant attention in recent years. This has been driven by the pressing need for ethical, trustworthy, and transparent implementation of AI systems in healthcare. The following review provides a concise overview of REF-AI in the context of medical image analysis. It begins with the fundamental concepts of AI responsibility, explainability, and fairness, followed by a comprehensive taxonomy of over 35 key algorithms and strategies. In addition, it compares methodologies, strengths, and limitations, such as the alignment of AI models with medical standards and the development of interpretable and actionable results for clinicians. Finally, it highlights current trends and proposes directions for future research to further advance the responsible, explainable, and fair application of AI in medical imaging.