[Artificial intelligence in radiology practice : between dream and reality].

Authors

Abstract

Despite its rapid development and promising performance reported in experimental settings, artificial intelligence (AI) in radiology still faces major limitations, particularly in terms of generalizability, explainability, and integration into clinical practice. At the same time, expectations among radiology professionals remain high, with AI perceived as a potential lever to improve diagnostic quality, optimize workflows, and free up time for higher-value clinical tasks. Although many promises have yet to be realized, several applications have already demonstrated measurable clinical benefits and are being gradually integrated into imaging devices.

Tags

Topics