The Rise of Specialized 'Agentic AI'
The AI landscape is shifting from single large models to federated systems of smaller, specialized AIs managed by an orchestration layer, a recent podcast noted. These "agentic AIs" can autonomously execute complex, multi-step goals, creating a need for new roles like "AI workflow controllers" to manage and monitor them in business operations.
- The U.S. Food and Drug Administration (FDA) has authorized over 1,300 AI-driven medical products, with nearly 80% (1,039) specifically for radiology as of early 2026. This represents a significant increase from just over 500 total approved devices at the beginning of 2023, with major manufacturers like GE HealthCare, Siemens Healthineers, and Philips leading in clearances. - Agentic AI is moving beyond just flagging findings to automating the entire follow-up process. These systems can identify patients needing follow-up, prepare pre-authorization documents, and coordinate scheduling, aiming to create a safety net that prevents patients from being lost to follow-up. - Major equipment manufacturers are embedding agentic AI and orchestration layers directly into their imaging platforms. For instance, GE HealthCare's Imaging 360 uses AI to help balance device utilization and standardize protocols, while Siemens Healthineers' AI-Rad Companion provides AI-powered workflow solutions to reduce repetitive tasks. - The shift from fee-for-service to value-based care, along with changes in reimbursement, is accelerating the movement of imaging procedures out of hospitals and into freestanding outpatient centers. Approximately 40% of all radiology volume is now performed in outpatient settings, a trend driven by lower costs and greater patient convenience. - The Centers for Medicare & Medicaid Services (CMS) has finalized a 2.83% reduction in the Medicare Physician Fee Schedule conversion factor for 2025, continuing a trend of declining reimbursement that impacts imaging providers. However, in a significant policy change, CMS will now separately pay for diagnostic radiopharmaceuticals with a per-day cost exceeding $630, which was previously bundled. - The demand for imaging services is rising while the U.S. faces a projected shortfall of up to 42,000 radiologists by 2033. AI is seen as a critical tool to manage this gap by automating tasks like image processing and triage, which can help alleviate radiologist burnout and improve efficiency. - AI orchestration platforms act as an intelligent control layer, ensuring the correct imaging study is routed to the right specialist and appropriate AI algorithms are applied, regardless of where the image was acquired. This approach helps to overcome data silos between different departments, such as radiology and cardiology, allowing for a more holistic view of the patient. - The next frontier for agentic AI involves autonomous robotics in imaging. Collaborations, such as the one between NVIDIA and GE HealthCare, are focused on developing autonomous ultrasound and X-ray systems that can automate patient positioning, image scanning, and quality checks, potentially enabling use in remote or minimally staffed locations.
