FDA creates transparency principles for AI in medical devices

The Food and Drug Administration (FDA), alongside Canada and the U.K., has established guiding principles for transparency in machine learning-enabled medical devices. These principles ensure crucial information affecting risks and patient outcomes is clearly communicated.

In 2021, the FDA, Health Canada, and the U.K.'s Medicines and Healthcare products Regulatory Agency (MHRA) jointly developed 10 principles for good machine learning practices. These focused on transparency, advocating for clear, essential information to users, though specifics were initially sparse.

Recently, the agencies released a follow-up document defining transparency as the clarity of information about machine learning devices communicated to stakeholders. It emphasizes providing explanations for algorithm outputs like diagnostic recommendations, enabling users to critically assess devices during patient care.

The document advises on what, how, and when information should be shared, urging optimized user interfaces for effective communication and timely updates on device performance and changes.