The U.S. Food and Drug Administration (FDA) released a paper Oct. 27, of 10 guiding principles for good machine learning practice (GMLP) related to medical device development. The principles were developed by the FDA with counterpart Canadian and United Kingdom regulatory agencies.
The 10 guiding principles are:
- Multi-Disciplinary Expertise Is Leveraged Throughout the Total Product Life Cycle.
- Good Software Engineering and Security Practices Are Implemented
- Clinical Study Participants and Data Sets Are Representative of the Intended Patient Population.
- Training Data Sets Are Independent of Test Sets.
- Selected Reference Datasets Are Based Upon Best Available Methods.
- Model Design Is Tailored to the Available Data and Reflects the Intended Use of the Device.
- Focus Is Placed on the Performance of the Human-AI Team.
- Testing Demonstrates Device Performance during Clinically Relevant Conditions.
- Users Are Provided Clear, Essential Information.
- Deployed Models Are Monitored for Performance and Re-training Risks are Managed.
The release of the principles comes in the wake of the FDA’s Oct. 14 public workshop about artificial intelligence (AI) transparency, which featured an American College of Radiology® (ACR®) presentation and panel participation.
For more information about ACR advocacy related to FDA regulatory oversight of AI and machine learning-based radiology software, please contact Michael Peters, ACR Director of Government Affairs.