AI in medical devices under MDR and IVDR regulations

Artificial Intelligence in Medical Devices:
Insights from the German Notified Bodies

In the rapidly evolving landscape of medical technology, Artificial Intelligence (AI) is emerging as a transformative force. To address the unique challenges and opportunities AI presents, the German Notified Bodies, in collaboration with Team-NB, have developed a comprehensive position paper and questionnaire aimed at guiding manufacturers through the complexities of AI integration in medical devices.

Key Highlights from the Position Paper

The joint publication, titled Artificial Intelligence in Medical Devices, emphasizes the importance of a process-oriented approach to ensure the safety and efficacy of AI-based medical devices. Unlike traditional software, AI systems often exhibit adaptive and autonomous behaviors, which necessitate additional scrutiny during their lifecycle.

Why This Guidance Matters

Under the European Medical Device Regulation (MDR) and In Vitro Diagnostics Regulation (IVDR), manufacturers are required to demonstrate compliance with stringent safety and performance standards. However, AI-specific challenges—such as self-learning mechanisms and stochastic modeling—demand tailored approaches. This document provides actionable guidance for:

  1. Development and Risk Management:
  • Identifying roles and competencies specific to AI projects.
  • Documenting AI-related risks and ensuring compliance with regulatory requirements.
  1. Data Management:
  • Ensuring data integrity during training, validation, and testing phases.
  • Addressing biases and ensuring datasets are representative of target populations.
 
  1. Product Validation and Post-Market Surveillance:
  • Establishing robust validation processes for AI models.
  • Preparing comprehensive post-market surveillance plans to monitor real-world performance.

AI-Specific Challenges and Solutions

One of the standout aspects of this guidance is its focus on cybersecurity risks unique to AI, including adversarial attacks. It underscores the need for manufacturers to incorporate robust cybersecurity measures and risk mitigation strategies, aligning with the AI Regulation (EU) 2024/1689.

Collaboration and Continuous Improvement

The publication also highlights the role of collaboration between manufacturers, regulators, and notified bodies. By following the best practices outlined, manufacturers can not only meet compliance requirements but also set a benchmark for innovation and safety in medical AI.

Conclusion

The German Notified Bodies and Team-NB’s Artificial Intelligence in Medical Devices position paper is an essential resource for navigating the complexities of AI integration. By adhering to this guidance, manufacturers can confidently bring innovative and compliant AI-based medical devices to market.

For more detailed insights and to download the questionnaire, visit Team-NB’s official website or the German Notified Bodies portal.