Two clearances, two corners of imaging AI
On November 4, 2026, the FDA granted 510(k) clearance to the Evident Vascular Guidance System and its peripheral imaging catheters, built by San Jose-based Evident Vascular. It's an intravascular ultrasound (IVUS) platform used during peripheral vascular interventions — the catheter goes inside the vessel during a procedure, pairing custom imaging electronics with NVIDIA-accelerated real-time processing, flush-less plug-and-play catheters, and hands-free voice control. Eric A. Secemsky, MD, Director of Vascular Intervention at Beth Israel Deaconess Medical Center, said in the company's release that "the vessel-edge clarity and consistency across the field of view are remarkable," adding that the imaging helps physicians measure vessels, plan therapy, and assess procedural results.
A few weeks earlier, Sarasota-based Deep Reasoning AI received 510(k) clearance for NeuroDR, software that runs automatically once a head-and-neck CT angiography (CTA) scan finishes. It strips out bone, soft tissue, and veins, then generates 3D maximum intensity projections, volume renderings, and curved planar reformats of the carotid and vertebral arteries and sends them to PACS as DICOM images. Both are real, useful, narrowly scoped tools — one for a cath lab procedure, one for a single vascular-visualization step. Neither drafts a diagnostic report.
The bigger number behind both clearances
Zoom out and these two clearances are drops in a fast-filling bucket. The FDA's own AI-enabled medical device page states the agency has authorized more than 1,600 such devices for marketing in the US as of September 2026. An independent analysis of that public list found 335 new authorizations in 2025 alone — a 43% jump over 2024's 235 — up from just 6 authorizations in 2015, pushing the cumulative count to 1,433 devices by the end of 2025. Through June 29, 2026, the same tracker counted 181 more.
Imaging drives most of that growth, and radiology specifically is the biggest single category. One tracking analysis found that of the roughly 1,356 AI devices the FDA had authorized through September 2025, 1,039 — about 77% — were radiology devices. Almost all of these clearances, across every modality, go through the lower-friction 510(k) pathway rather than the more rigorous De Novo or premarket-approval routes, by showing the new device is substantially equivalent to one already on the market.
What this bundle of clearances is — and what it isn't
| Clearance | Modality / setting | What it outputs | Scope |
|---|---|---|---|
| Evident Vascular Guidance System | IVUS catheter, cath lab | Live vessel images during the procedure | Single interventional step |
| Deep Reasoning AI NeuroDR | CTA, head & neck | 3D renderings sent to PACS | Single visualization step |
| Comprehensive CT reporting workflows | CT, reading room | Structured report draft, every finding | Full study, end-to-end |
That's the pattern worth noticing: the vast majority of the FDA's 1,600-plus AI clearances, including these two, cover a single task on a single piece of anatomy — detect this, render that, enhance this image. Very few cover a complete diagnostic report across an entire study. A high clearance count is a real signal of regulatory momentum in imaging AI broadly, but it isn't a proxy for how much of the actual reporting workload any single tool addresses.
Where this matters for radiology groups and imaging centers
For buyers evaluating AI, the practical takeaway is to ask what a clearance actually covers before assuming it solves a reporting bottleneck. A 510(k) for a visualization or acquisition tool says nothing about whether a system can produce a comprehensive, ready-to-sign CT report draft across the full range of findings a radiologist has to address — that's a different category of product, built around covering an entire study rather than one task within it, with a radiologist reviewing the draft before the reading radiologist signs. As the FDA's AI device count keeps climbing, distinguishing "point tool" clearances from full-study reporting tools will matter more, not less, to anyone building a vendor shortlist.
Frequently asked questions
What did the FDA clear for intravascular ultrasound in November 2026?
On November 4, 2026, the FDA granted 510(k) clearance to the Evident Vascular Guidance System and Evident peripheral imaging catheters, made by San Jose-based Evident Vascular. The system is intended for intravascular ultrasound (IVUS) examination during peripheral vascular interventions, pairing high-fidelity imaging with NVIDIA-accelerated real-time processing, flush-less plug-and-play catheters, and hands-free voice control.
What does Deep Reasoning AI's NeuroDR software do?
NeuroDR, cleared by the FDA in October 2026, is CT angiography software that automatically visualizes the arteries of the head and neck once a scan is complete. It removes bone, soft tissue, and veins from the dataset, then sends 3D maximum intensity projections, volume renderings, and curved planar reformats of the carotid and vertebral arteries to PACS as DICOM images.
How many AI-enabled medical devices has the FDA authorized overall?
The FDA states it has authorized more than 1,600 AI-enabled medical devices for marketing in the US as of September 2026. Independent analysis of the FDA's public list shows 335 new authorizations in 2025 alone, up from just 6 in 2015, pushing the cumulative count to 1,433 devices by the end of that year.
Why does radiology dominate FDA AI device clearances?
Imaging is data-rich, standardized, and well-suited to pattern recognition, and most imaging AI clearances go through the lower-friction 510(k) pathway by showing substantial equivalence to an existing device. One industry tracking analysis found radiology accounted for about 77% of the roughly 1,356 AI devices the FDA had authorized through September 2025.
Source: Radiology Business, Vascular News, AuntMinnie, the FDA's AI-Enabled Medical Devices list, independent tracking by aiadoption.org, and The Imaging Wire. Figures are rounded as reported.