Over the past decade, the number of medical devices incorporating artificial intelligence (AI) has seen a significant surge. As of August 7, 2024, the U.S. Food and Drug Administration (FDA) had authorized nearly 1,000 AI or machine learning-enabled devices, with the pace of approvals sharply increasing in recent years. While the FDA approved its first AI-powered device in 1995, the trend has accelerated, particularly from 2015 onward, when just six AI devices were authorized. By 2023, that number had surged to 221, and in 2024, the FDA had already cleared 107 devices, putting the industry on track to match or exceed last year’s figures.
Experts attribute this rapid growth to several factors: a greater focus on connected devices, increasing investments in AI, and an evolving understanding of how to regulate software as medical devices. “We’re definitely seeing huge increases in investment. There’s no doubt about that,” said Jennifer Goldsack, CEO of the Digital Medicine Society, underscoring the industry’s growing confidence in AI’s potential to enhance patient care.
AI in Medical Devices: Key Innovations
AI’s integration into medical devices spans various specialties, with a major focus on imaging, diagnostics, and treatment planning. Some notable examples include:
- AI-Rad Companion by Siemens Healthineers: This tool enhances clinical image analysis, providing both qualitative and quantitative insights to aid in diagnostics.
- LumineticsCore by Digital Diagnostics: An AI system capable of automatically diagnosing diabetic retinopathy from eye images, eliminating the need for a specialist’s interpretation.
- Atrial Fibrillation History Feature by Apple: Integrated into the Apple Watch, this feature tracks and reports incidents of atrial fibrillation using real-time data, marking a significant step in wearable health tech.
Established medtech giants like GE Healthcare, Siemens Healthineers, and Medtronic are leading the charge, developing AI-powered imaging and diagnostic tools. At the same time, startups such as Aidoc, RapidAI, and Butterfly Network are introducing innovative solutions for ultrasound imaging and condition detection. Even companies outside traditional healthcare, like Apple and Nvidia, are contributing AI expertise to the medical device sector, signaling a growing trend of cross-industry collaboration.
A Closer Look at Specialties: Radiology and Beyond
As of August 2024, the vast majority of FDA-authorized AI devices are used in radiology, where AI is particularly effective at improving image quality, optimizing scan times, and aiding diagnosis. For example, more than three-quarters of AI devices authorized to date are radiological tools, including AI-driven triage software and radiation therapy planning systems.
However, AI’s applications are expanding into other medical fields, particularly cardiology. With 98 AI-enabled heart-related devices approved by the FDA, AI is being used to analyze electrocardiograms and detect conditions like arrhythmias or heart failure. The cardiovascular sector is a key area where AI is showing promise, and experts like Goldsack are eager to see how these technologies evolve in specialties that have less experience with AI.
Robotics is also emerging as a new frontier for AI in healthcare. AI tools are being used to assist in preoperative planning, intraoperative visualization, and post-surgery analytics. According to Nvidia’s David Niewolny, “What started in the radiology suite is now migrating to other areas of the hospital.”
Market Leaders and Future Directions
The leaders in AI medical devices are GE Healthcare and Siemens Healthineers. GE Healthcare, with 81 AI devices authorized as of August 2024, has developed cutting-edge tools like Air Recon DL, which improves MRI scan efficiency by reducing scan times by up to 50%. Similarly, Siemens Healthineers, with 70 FDA-cleared AI devices, focuses on both diagnostic algorithms and imaging tools, with notable products like AI-driven radiation therapy planning software.
Both companies are increasingly integrating AI into their offerings, moving beyond standalone devices to include AI-powered services, subscription models, and enterprise-level systems. Looking to the future, companies like Siemens are particularly interested in multimodal AI, which combines imaging data, lab results, and patient histories to aid clinical decision-making.
Regulatory Pathways and Challenges
A large majority (about 97%) of AI medical devices are cleared through the FDA’s 510(k) pathway, which allows for faster and less rigorous approvals for devices deemed “substantially equivalent” to existing products. Fewer devices undergo the more stringent premarket approval (PMA) process, which is reserved for high-risk devices.
However, despite the rapid growth, the regulatory landscape remains complex. AI tools that suggest treatment decisions or diagnoses are subject to FDA oversight, while tools that assist with administrative tasks, like scheduling or inventory management, are generally exempt. The FDA’s evolving guidelines, particularly for software that provides diagnostic recommendations, are shaping the future of AI regulation in healthcare.
In conclusion, the surge in AI-enabled medical devices represents a significant shift in the healthcare landscape, with major opportunities for innovation across various specialties. As more devices are authorized and technology advances, the integration of AI into healthcare is expected to continue expanding, potentially transforming diagnostics, treatment, and patient care. However, as the industry grows, regulatory bodies will need to keep pace to ensure patient safety while fostering innovation.
Source: BioPharmaDive
