Answers are grounded in this paper and cite their sources.
Background: Artificial intelligence is used to detect and characterize abnormalities in diagnostic imaging and to support clinical interpretation. Technical performance does not always translate into patient benefit. Aim: To evaluate applications of artificial intelligence (AI) in diagnostic imaging, focusing on diagnostic performance, collaboration with clinicians, and limitations of clinical implementation. Methods: To identify recent articles on the use of artificial intelligence in diagnostic imaging, the PubMed database was searched. Only English-language publications were included. Results: AI algorithms have promising capabilities for detecting and characterizing abnormalities in mammography, radiography, computed tomography, magnetic resonance imaging, and ultrasonography. In selected settings, they improve sensitivity, support segmentation, and reduce interpretation workload. However, they do not consistently improve specificity, time to intervention, or patient outcomes. The clinical evidence varies in methodological quality and representativeness. Conclusions: AI should primarily be considered a decision-support technology rather than an autonomous replacement for clinical judgment. Wider adoption requires external validation, workflow-based evaluation, and ongoing safety monitoring.
No takes yet. Share an insight, caveat, or question.
Czerwonka et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: