Back to all papers

Regulatory frameworks and liability for AI-enabled radiology devices: Analysis of U.S. and European Union approaches.

November 13, 2026pubmed logopapers

Authors

Shabanan SH,Kutlu AZ,Mezrich JL

Affiliations (3)

  • Yale School of Medicine, Department of Diagnostic Radiology and Biomedical Imaging, 333 Cedar Street, TE2, New Haven, CT, 06520, United States of America. Electronic address: [email protected].
  • Yale School of Medicine, Department of Diagnostic Radiology and Biomedical Imaging, 333 Cedar Street, TE2, New Haven, CT, 06520, United States of America. Electronic address: [email protected].
  • Yale School of Medicine, Department of Diagnostic Radiology and Biomedical Imaging, 333 Cedar Street, TE2, New Haven, CT, 06520, United States of America. Electronic address: [email protected].

Abstract

Rapid expansion of artificial intelligence (AI) in medical imaging demands regulatory clarity and robust liability frameworks to ensure safety and public trust. This study describes clinical studies identified through a targeted ClinicalTrials.gov search relevant to AI-enabled radiology devices, and discusses how U.S. and European Union regulatory and liability frameworks may influence transparency and accountability. FDA AI/ML radiology device listings were filtered for radiology entries to assemble a descriptive inventory (n = 1094, Accessed April 9, 2026). gov was searched using four AI-in-imaging terms and deduplicated by NCT identifier (n = 153 unique records). Study status, type, sponsor category, imaging modality, clinical purpose, and results-posting status were summarized descriptively. This limited search was intended to be illustrative rather than a comprehensive estimate of clinical validation. Among the 153 identified records, sponsorship was predominantly academic (68.0%), with smaller shares from government (13.7%), industry (8.5%), individual investigators (7.2%), and nonprofit organizations (2.6%). Most records were observational (75.8%) rather than interventional (24.2%). No trial had posted results, including among the 34 completed studies with a primary completion date more than 12 months before the search date. Analysis did not establish whether these records were subject to mandatory results reporting requirements. A separate FDA inventory contained 1094 radiology entries; No device-level linkage to registry records was performed. This targeted-search registry snapshot provides context for a broader discussion of regulatory lifecycle oversight, documentation and monitoring practices for AI device liability under U.S. and EU frameworks. Clearer device-trial linkage, standardized reporting and updated liability frameworks may improve accountability and legal clarity.

Topics

Journal Article

Ready to Sharpen Your Edge?

Subscribe to join 11k+ peers who rely on RadAISlice. Get the essential weekly briefing that empowers you to navigate the future of radiology.

We respect your privacy. Unsubscribe at any time.