The European Society of Radiology (ESR) AI Working Group has finalized and published its statement: “Guiding AI in Radiology: ESR’s Recommendations for Effective Implementation of the European AI Act.”
As the European AI Act is set to shape the future of AI regulation in healthcare, ESR’s this statement provides an in-depth analysis of its impact on radiology and offers clear, actionable recommendations to ensure safe, reliable, and patient-centered AI integration in medical imaging.
About the Statement
Developed through collaboration between radiologists of the ESR AI Working Group, the statement identifies potential gaps and uncertainties in the AI Act and addresses nine key areas that are crucial for radiology:
- AI Literacy – Strengthening education and training for radiologists and healthcare professionals.
- High-Risk Classification – Clarifying the classification of AI tools under medical device regulations.
- Data Governance – Ensuring high-quality, unbiased, and representative imaging datasets.
- Transparency & Information Sharing – Requiring AI providers to disclose system details to users.
- Human Oversight – Safeguarding clinical decision-making and preventing automation bias.
- Quality Management Systems – Aligning compliance between the AI Act and the Medical Device Regulation (MDR).
- Obligations of AI Deployers – Defining clear responsibilities for radiologists using AI tools.
- Regulatory Sandboxes – Evaluating AI models in controlled environments before clinical integration.
- Post-Market Monitoring & Surveillance – Developing robust monitoring systems to track AI performance over time.
Why This Matters
With the AI Act entering its implementation phase, it is essential that radiologists contribute to the shaping of how AI is regulated in clinical practice. ESR is committed to working with EU policymakers to ensure that regulation supports innovation while maintaining the highest standards of patient safety and clinical effectiveness.
Read the full statement here: https://insightsimaging.springeropen.com/articles/10.1186/s13244-025-01905-x