The FDA, that recently released its Digital Health action plan, and will take a multi-pronged approach to the development of safe software based on Artificial Intelligence and machine learning.
The document published on January 12 on the official FDA (Food and Drug Administration) website, takes into account five central axes of action for software as a medical device (SaMD) to continue their development and be applied. This essentially contemplates those based on Artificial Intelligence (AI) and machine learning:
- Continue the development of the proposed regulatory framework, including by issuing draft guidance on a pre-determined change control plan (for software learning over time);
- Support the development of machine learning best practices to evaluate and improve machine learning algorithms;
- Encourage a patient-centered approach, including device transparency to users;
- Develop methods to evaluate and improve machine learning algorithms; and
- Continue pilot testing of real-world performance monitoring.
On the first point, the FDA, plans to publish guidance later this year that addresses more elements of safety and efficacy of SaMDs. And also, to promote regulations with greater personalization in the face of new technologies.
Also, the FDA in its efforts to seek as much transparency as possible regarding SaMDs, seeks the implementation of a public workshop on how labeling on devices improves trust in devices, especially those based on AI and machine learning. “Promoting transparency is a key aspect of a patient-centered approach, and we believe this is especially important for AI/ML-based medical devices, which may learn and change over time, and which may incorporate algorithms exhibiting a degree of opacity,” the FDA explained.
Similarly, the agency seeks to reduce opacity in the performance of AI and machine learning algorithms through a new methodology for their evaluation. In this sense, enters the last of the five points, which points to the real world, “Gathering performance data on the real-world use of the SaMD may allow manufacturers to understand how their products are being used, identify opportunities for improvements, and respond proactively to safety or usability concerns. Real-world data collection and monitoring is an important mechanism that manufacturers can leverage to mitigate the risk involved with AI/ML-based SaMD modifications, in support of the benefit-risk profile in the assessment of a particular marketing submission,” explained the agency.
