A call for clinical evaluation of open-source algorithms in radiology

Artificial intelligence (AI) is rapidly transforming numerous fields, and radiology stands at the forefront of this revolution [1]. AI-related research now dominates radiology publications, with every major journal issue featuring new algorithms promising improvements in lesion detection, segmentation, or classification. This enthusiasm is mirrored by growing clinical adoption: approximately 48% of European radiologists report using AI, mainly in computed tomography (CT) and magnetic resonance imaging (MRI) [2].