FDA Publishes AI/ML-Based Software as a Medical Device Action Plan

The U.S. Food and Drug Administration (FDA) has published its first Artificial Intelligence/Machine Learning (AI/ML)-Based Software as a Medical Device (SaMD) Action Plan. According to an agency news release, the blueprint outlines a detailed approach to furthering the FDA’s oversight of AI/ML-based medical software. 

Drafted by the Agency’s Digital Health Center of Excellence (DHCE), the SaMD Action Plan responds to the following concerns that emerged from its 2019 request for feedback on the evolving technology:

• Expanding proposed regulatory framework, including guidance on the “Predetermined Change Control Plan” which accounts for the methodology used to implement changes in a controlled manner that manages risks to patients;
• Supporting thoughtful machine learning practices that evaluate and improve algorithms;
• Fostering a patient-centered approach and promoting device transparency;
• Developing tools to assess and improve algorithms; and 
• Advancing performance monitoring mechanisms. 

The framework, explains DHCE Director Bakul Patel, will advance “the enormous potential that these technologies have to improve patient care while delivering safe and effective software functionality that improves the quality of care that patients receive.”

The FDA says it currently receives a high number of marketing submissions and pre-submissions for AI/ML products, and it anticipates the volume will only continue to grow. The agency intends to tailor the SaMD Action Plan accordingly as technology progresses.

Given the expanding role of AI in healthcare, MLMIC policyholders are advised to monitor all FDA developments related to the application of AI/ML-based medical software. Additionally, MLMIC recommends any AI platforms that are in current use should be routinely monitored for updates and efficacy.