RadioVal was featured at the European Congress of Radiology (ECR) 2025, demonstrating the transformative potential of AI in clinical decision-making. Presented as part of the session “Intelligence at Work: Harnessing Artificial Intelligence (AI) for Clinical Decision-Making,” the project was showcased alongside AI-POD and ODELIA. This session, held on 1 March 2025 from 16:00 to 17:30 CET, sparked insightful discussions on the evolving role of AI in healthcare. The session was led by two distinguished experts: Prof. Karim Lekadir from the University of Barcelona and Gianna Tsakou from Gruppo Maggioli.
Karim Lekadir unveiled the FUTURE-AI guideline, a comprehensive framework designed to ensure AI tools in healthcare are trustworthy and effective. Drawing on a survey of clinicians, Lekadir revealed that over 70% of healthcare professionals have not yet used AI tools in their practice. This statistic underscores the need for guidelines that promote confidence in AI adoption. The FUTURE-AI paper identifies key characteristics of trustworthy AI and outlines best practices for operationalizing these technologies, emphasizing that trustworthiness is a continuous process that requires ongoing evaluation and feedback. Multi-stakeholder engagement, including input from clinicians, researchers, and policymakers, is critical for developing AI tools that can be fully integrated into healthcare systems.
Ioanna Tsakou shared her insights on the development of the AI Passport, a tool created by RadioVal to ensure standardized AI descriptions and traceability. The AI Passport is a web-based platform that collects and organizes traces related to the data sets, AI models, evaluations, usage, and maintenance of AI tools. By offering a mechanism for comprehensive monitoring, the AI Passport enhances the trustworthiness of AI tools, providing essential input for quality control and auditing. Tsakou highlighted that the tool’s adaptability makes it suitable for any medical AI, ultimately supporting transparent and accountable AI implementation.
Both experts stressed the importance of AI traceability, with Tsakou emphasizing that the ability to track and validate AI tools’ performance builds trust and encourages widespread adoption. Lekadir’s call for multi-faceted AI evaluation further reinforces the notion that building trust in AI requires rigorous, ongoing efforts at every stage of its lifecycle.
Open access for all attendees:
- Visit the session page on ESR Connect and select the session you would like to attend.
- Click on ‘Login’ in the upper right corner and enter your ESR MyUserArea credentials or create a new ESR account for free (after you have created your account please visit the session page once again).
- The session is now saved in your user menu under ‘My Courses’ which can be accessed via your user icon in the upper right corner.
- When the session starts, log-into your account and visit the session page, where the livestream will start automatically.
