Integrating AI with care: FDA reveals focus areas for forthcoming guidance on AI products 

March 25, 2024

AI is everywhere. Its applications are endless, and they continue to expand as more people adopt and become comfortable with the technology. Within the healthcare sphere, AI has the potential to increase speed and efficiency of diagnoses, increase patient engagement and decrease nonconformance, and help with administrative functions. While AI is sure to improve many aspects of the medical industry, it is also important to integrate AI with care to ensure safe and effective use of devices with AI.

In light of AI’s growth within healthcare, the FDA recently published a white paper outlining the four focus areas that the Center for Biologics Evaluation and Research (CBER), Center for Drug Evaluation and Research (CDER), Center for Devices and Radiological Health (CDRH), and Office of Combination Products (OCP) will be considering as they move forward to create joint guidance for products utilizing AI. The four main focus areas were as follows:

Foster collaboration to safeguard public health. The FDA’s medical product Centers collaborate extensively with various stakeholders, including developers, patient groups, academia, and global regulators. Their goal is to foster a patient-centered regulatory approach that emphasizes collaboration and health equity.

Advance the development of regulatory approaches that support innovation. The FDA’s medical product Centers are committed to ensuring regulatory consistency across Centers for the use of AI in advancing public health and innovation.

Promote the development of standards, guidelines, best practices, and tools for the medical product lifecycle. The FDA’s medical product Centers are committed to ensuring safety and effectiveness standards for AI-enabled medical products. They plan to refine and develop considerations over time.

Support research related to the evaluation and monitoring of AI performance. The FDA’s medical product Centers seek to support projects that demonstrate how bias may be introduced and how health inequities may exist in data and models.

While guidance may not be written yet, there are specific throughlines across existing FDA guidance that can be utilized as a source of direction and clarity:

  • Safe and effective use for users is of primary importance. Ultimately, no matter the product, the FDA prioritizes safe and effective use for users. Minimizing use-related risk to the lowest possible level is critical for any product.
  • Generating AI models should be completed with data reflective of the end user. Just like any medical product should be validated with populations that use it, AI and machine learning programs should be geared toward their users. Follow the FDA’s Good machine learning practices to aid in model development. Remember: garbage in means garbage out. In other words, a model is only as good as the data that was used to develop it.
  • Communication and collaboration is ongoing. The FDA Centers are actively working to better understand medical products that incorporate AI and are seeking to learn from subject matter experts. If you have a medical product that incorporates AI, it’s likely they’ll want to learn from you.

In the near future, special attention will be given by the FDA to medical products incorporating AI. This means that it’s important to consider how to adjust use-related risk analyses and formative and summative evaluations to account for AI.

Upfront planning and consideration of the human factors elements of product development in relationship to AI will lead to less surprises during the process and support teams to achieve optimal user-centered outcomes.

Even though the AI landscape is newer to medical devices, it’s been used in consumer and tech applications for some time. We have worked with multiple clients on numerous studies with products utilizing AI. This, coupled with our expertise in all steps of the medical device product development life cycle, can help guide you through your process.

The AI landscape is ever evolving, and FDA regulatory approaches are evolving along with it. The right human factors research partner can guide your AI product development through this changing landscape–and Bold Insight has the experience to help.

If you’re interested in starting a conversation, reach out! We’d love to hear from you.