First REALM Webinar on AI enhanced Medical Device Software Evaluation attracts regulators and healthcare professionals

Online, 25 March 2025 – The REALM project consortium hosted a public webinar to showcase how real-world data and AI evaluation tools are set to reshape the regulatory landscape for the evaluation of Medical Device Software across Europe. The event attracted a diverse group of stakeholders, including regulators, health technology assessors, software developers, and academic researchers.

Funded under the Horizon Europe framework, the REALM project (Real-world data Enabled Assessment for regulatory decision-making of medical device software) aims to develop a collaborative framework where stakeholders can jointly create and evaluate innovative Medical Device Software, directly benefiting patients and healthcare professionals .

“This four-year research and innovation project focuses on real-world data-enabled assessment for regulatory decision-making,” said Janaki Raman Rangarajan from the Virtual Physiological Human Institute. “We’re building a transparent and robust environment for AI tools to be fairly assessed and safely brought to market.”

A New Regulatory Framework

REALM’s approach combines a federated data infrastructure, synthetic data generation, a regulatory toolbox, model cards, and post-market surveillance mechanisms — all integrated into a collaborative platform known as the REALM Sandbox.

“We’re bridging the gap between rapid healthcare innovation and the slower pace of regulation,” explained Dr Gökhan Ertaylan from the Flemish Institute for Technological Research (VITO). “REALM proposes a harmonized European approach through cross-border cooperation and regulatory upskilling.”

Putting the Tools in Stakeholders' Hands

The event highlighted the launch of REALM Academy, a structured training program mainly targeting regulators, AI developers, and health technology assessment (HTA) bodies.

“REALM Academy turns all these tools into something practical,” said Dr Anshu Ankolekar from Maastricht University. “Evaluation isn’t just technical — it’s a collaborative process. Each stakeholder comes with different questions, and REALM is trying to address them all.”

REALM Academy will launch its first round of online training later this year, accompanied by hands-on workshops. The consortium invites stakeholders from across the healthcare innovation ecosystem to get involved. Greater engagement will lead to a better alignment between current challenges in evaluation and the solutions REALM offers! More information about the REALM Academy is available here.

Use Cases in Action

The platform already supports five real-world use cases involving AI in imaging, genomics, hospital readmission prediction, and intensive care. One example is the DuneAI model for lung cancer segmentation, evaluated using real and synthetic data.

“With over 7,000 AI models proposed in healthcare last year alone, we urgently need frameworks that verify claims and protect patients,” added Dr Chang Sun from Maastricht University. “REALM does this transparently, without compromising data privacy.”

Next Steps

The first in-person workshop will take place on July 7th, 2025, in Brussels, offering a hands-on look at the REALM platform and toolbox. Additional in-person workshops, webinars, and online training tracks will follow throughout 2025–2026.

“We’re not just building a platform — we’re co-creating a future-proof process for safe and timely AI integration into European healthcare,” said Janaki in his closing remarks.

Watch Tthe full recording of the webinar is available here:

For more information, visit: www.realm-project.eu

Follow REALM on LinkedIn for news, training updates, and upcoming events.