November 19, 2024 | 1 minute read

Upcoming FDA Meeting Focuses on Generative AI in Medical Devices

Get in touch

Jefferson Lin

Associate

Get in touch

Jefferson Lin

Associate

The U.S. Food and Drug Administration will focus on “Total Product Lifecycle Considerations for Generative AI-Enabled Devices” during its upcoming Digital Health Advisory Committee meeting on November 20–21. This meeting will address the unique challenges and opportunities presented by the use of generative AI in medical devices.

What are Generative AI-Enabled Devices?

Generative AI (GenAI) refers to AI models that can create new content, like images, text, audio, and video, by learning patterns from existing data. GenAI-enabled devices are medical devices that incorporate GenAI methods or models as essential parts of their functionality. The FDA emphasizes that GenAI-enabled devices are not a single category, and considerations for one device may not apply to all.

Importance of a Total Product Lifecycle Approach

GenAI-enabled devices often rely on models designed to evolve rapidly over time. This underscores the importance of a Total Product Lifecycle approach, which considers the device's safety and effectiveness throughout its entire lifespan, from design and development to post-market use.

Key Challenges and Discussion Questions

The FDA's Executive Summary for the meeting highlights several challenges associated with GenAI-enabled devices:

Rapid evolution of models: GenAI models are constantly changing, making it difficult to evaluate their long-term safety and effectiveness.
Limited information about training data: Details about the data used to train foundation models may be unavailable, making it harder to assess potential biases and limitations.
Hallucinations: GenAI models can generate inaccurate or misleading outputs, which pose a risk to patient safety.
Usability risks: GenAI introduces new usability risks compared to non-generative AI, impacting healthcare professionals, patients, and caregivers.

The meeting will address these challenges through discussion questions focused on premarket performance evaluation, postmarket monitoring, and overall risk management strategies for GenAI-enabled devices.

The FDA is Asking for Expert Input:

What information is needed to evaluate the safety and effectiveness of GenAI-enabled devices premarket?
What evidence should be considered regarding performance evaluation and training data?
How can postmarket monitoring ensure devices maintain accuracy, relevance, and reliability?
What strategies can be implemented to manage performance and address potential biases across multiple deployment sites?

The novel capabilities of GenAI may offer unique benefits to patients and public health, but the use and adoption of GenAI also come with specific risks and complexities that challenge FDA’s approach to the regulation of devices. In particular, FDA faces challenges associated with applying a risk-based approach to classification and determining regulatory requirements for GenAI-enabled devices and, for those GenAI-enabled devices that require FDA’s regulatory oversight, FDA faces challenges associated with determining the types of valid scientific evidence for FDA’s evaluation of the safety and effectiveness of GenAI-enabled devices across the total product life cycle ("TPLC"). FDA has long promoted a TPLC approach to the oversight of medical devices, including AI-enabled devices, and has committed to advancing regulatory approaches for these devices.