AI/ML Medical Device Guidance
Korea has been at the forefront of AI/ML medical device regulation, with MFDS issuing initial guidance in 2017 and updated guidance in 2021 and 2024.
Core requirements (2024 guidance)โ
Pre-market requirementsโ
| Requirement | Details |
|---|---|
| Algorithm description | Full description of the model architecture, training methodology |
| Training data | Characteristics, size, representativeness, labelling quality |
| Validation data | Independent test dataset; performance metrics (sensitivity, specificity, AUC, etc.) |
| Clinical validation | Real-world performance in the intended use population |
| Explainability | For high-risk AI: explanation of how decisions are made |
| Intended use statement | Precise statement of what the AI is intended to do and in which setting |
Post-market requirementsโ
| Requirement | Details |
|---|---|
| Performance monitoring plan | Ongoing monitoring of real-world AI performance |
| Drift detection | Monitoring for algorithm performance degradation over time |
| Re-validation trigger | Define when re-validation is required (performance threshold breach) |
| Reporting | Integrate AI performance data into PMS and periodic safety reports |
Predetermined Change Control Plan (PCCP)โ
MFDS is developing PCCP guidance allowing manufacturers to pre-define the scope of future algorithm changes that do not require a new ํ๋ชฉํ๊ฐ submission.