For years, European imaging AI founders have been trapped by Rule 11 of the MDR, which effectively pushed almost all diagnostic software into Class IIa or higher. However, the regulatory landscape shifted dramatically this week as the European Commission’s simplification proposal moved toward implementation, signaling a fundamental win for the digital health sector.
Full Text of Rule 11
Software intended to provide information which is used to take decisions with diagnosis or therapeutic purposes is classified as class IIa, except if such decisions have an impact that may cause:
– death or an irreversible deterioration of a person’s state of health, in which case it is in class III; or
– a serious deterioration of a person’s state of health or a surgical intervention, in which case it is classified as class IIb.
Software intended to monitor physiological processes is classified as class IIa, except if it is intended for monitoring of vital physiological parameters, where the nature of variations of those parameters is such that it could result in immediate danger to the patient, in which case it is classified as class IIb.
All other software is classified as class I. [Source]
1. The End of the “IIa-by-Default” Nightmare
The core of the recent reform (Obelis (Jan 28, 2026)) addresses the “classification nightmare,” in which even low-risk software was subjected to the same scrutiny as invasive devices.
- The Change: Software used for monitoring, prediction, or diagnosis that does not carry a risk of death or serious deterioration may now be eligible for Class I.
- Impact: This allows for self-certification in many triage and screening applications, slashing time-to-market from 18 months to weeks and saving startups an average of €150k–€300k in initial audit fees.
2. Perpetual Certificates for Adaptive AI
Traditional MedTech regulation was built for hardware that doesn’t change. Software, particularly Adaptive AI, requires constant iteration.
- The proposal to replace the fixed 5-year certificate validity with a risk-based, continuous surveillance model means that software updates will no longer be held back by fear of triggering a full re-certification cycle.
3. The “Fast Lane”: Breakthrough Device Status (MDCG 2025-9)
Recent guidance (Johner Institute (Jan 21, 2026)) and analysis this week highlight the power of the Breakthrough Device (BtX) designation.
For imaging AI targeting life-threatening conditions (e.g., automated stroke detection or rare oncology), this status offers:
- Priority Review: Notified Bodies are now encouraged to prioritize these files.
- Scientific Advice: Direct access to EU Expert Panels to validate clinical evidence strategies before full submission.
4. AI Act vs. MDR: Clearer Boundaries
A critical clarification emerged this week (Osborne Clarke (Jan 28, 2026)): AI-based medical devices will likely be treated as lex specialis. High-risk AI requirements from the AI Act will be integrated into the MDR/IVDR process, ensuring that imaging companies face a single conformity assessment rather than duplicative audits.
Checklist: Is Your AI Ready for the “Breakthrough” Path?
Use this 5-point assessment based on the new MDCG 2025-9 framework:
- Criticality: Does your tool address a life-threatening or irreversibly debilitating condition?
- Innovation: Can you prove a high degree of technological novelty compared to the current “state of the art”?
- Clinical Impact: Is there a reasonable expectation of significant improvement in patient outcomes (e.g., faster time to treatment)?
- Evidence Lifecycle: Do you have a robust plan for Post-Market Clinical Follow-up (PMCF) to manage residual uncertainty?
- Regulatory Maturity: Is your Quality Management System (QMS) ready for the intense “structured dialogue” with Notified Bodies required for BtX status?