The FDA has authorized nearly 1,000 medical devices with artificial intelligence features between 1995 and August 2024, with a significant spike in approvals over the past decade. In 2015, the agency authorized six AI medical devices, compared to 221 devices in 2023. This surge is driven by increased investment in AI, growing familiarity with software regulation, and more connected devices.
Over three-quarters of authorized AI-enabled devices are in radiology, including features to improve image quality, optimize radiation dosing, and flag potential health conditions. However, companies are also developing AI-enabled devices in other areas, such as cardiovascular, neurology, and ophthalmology. Top companies with the most AI-enabled devices include GE Healthcare, Siemens Healthineers, and Canon.
The FDA has issued draft guidance on AI-enabled device software functions, emphasizing transparency, bias mitigation, and postmarket monitoring. The agency recommends a total product lifecycle approach to ensure safety and effectiveness of AI-enabled devices. Most AI devices are cleared through the 510(k) pathway, which requires demonstrating substantial equivalence to existing devices.
The FDA's approach aims to balance innovation with safety and effectiveness, enabling the development of new AI-enabled medical devices. Transparency and bias mitigation are crucial in AI-enabled devices, and manufacturers should have plans in place to monitor device performance and address changes or degradation over time. As the use of AI in medical devices continues to grow, the FDA's guidance will play a critical role in shaping the industry's approach to innovation and regulation.
