92 AI Devices in One Quarter. 75% of Them for Radiology.
The pace of AI-enabled medical device clearance in diagnostic imaging has reached a scale that is genuinely difficult to track from outside the field. The Imaging Wire’s June 2026 quarterly FDA analysis confirmed that in Q1 2026 alone, the FDA authorised 92 AI-enabled medical devices — 28% more than in Q4 2025, with 69 of those authorisations — 75% — for radiology devices. This is not a one-quarter anomaly. The proportion of AI device clearances going to radiology has remained consistently around 75% across quarters, reflecting the imaging specialty’s structural position as the clinical field where AI has the longest development history, the largest volume of training data, and the most clearly defined regulatory pathway through the 510(k) substantial equivalence route. The same analysis tracked the cumulative leader board: GE HealthCare leads with 130 total radiology AI authorisations (including recent acquisitions); Siemens Healthineers at 95; Philips at 58; Canon at 48; United Imaging at 40; and Aidoc — the leading pure-play radiology AI vendor — at 33.
What GE, Siemens, and Philips Are Actually Shipping in 2026
The clearance numbers are impressive, but the more instructive story is what specific products the major imaging equipment companies are commercialising in 2026 and how they are reshaping clinical workflows. In May 2026, Siemens Healthineers received FDA clearance for six new interventional imaging systems simultaneously — a single-announcement regulatory milestone that reflects both the company’s product portfolio breadth and the accelerating pace of its U.S. regulatory submissions. In the same month, the FDA cleared Siemens’ Mammomat B.brilliant system’s new contrast-enhanced mammography and CEM biopsy capabilities. In March 2026, Philips received FDA clearance for its Spectral CT Verida Family, an AI-powered detector-based spectral CT platform designed for radiology, cardiology, and oncology use cases — a product that Radiology Business described as part of a wave of 68 new FDA-cleared radiology algorithms in the preceding quarter. GE HealthCare, meanwhile, announced in February 2026 a $35 million expansion with BARDA to advance AI-powered ultrasound for trauma care and emergency preparedness — a government partnership that anchors AI ultrasound development in national emergency preparedness infrastructure rather than purely commercial deployment.
Aidoc’s 31 Clearances and 2,000 Hospitals: What Pure-Play AI Scale Looks Like
The most commercially instructive data point in the radiology AI market is not the total clearance count of any single manufacturer but the operational scale of the leading pure-play radiology AI company. Pinggy’s June 2026 analysis of AI medical imaging documented that Aidoc now holds more than 31 FDA-cleared tools and runs across nearly 2,000 hospitals, processing 60 million patient cases per year. In January 2026, Aidoc received FDA clearance for the first foundation model-powered clinical AI device — a single body CT triage solution covering 14 conditions simultaneously, including aortic dissection, appendicitis, bowel obstruction, and spleen injury. This multi-condition, single-model approach represents a meaningful departure from the single-condition point solutions that dominated the radiology AI market’s first generation: instead of deploying one algorithm for PE detection and another for intracranial haemorrhage, a foundation model processes the scan once and flags multiple potential findings simultaneously, with direct implications for workflow efficiency and radiologist productivity in emergency settings.
Bristol Myers Squibb and Microsoft: When Pharma Buys AI Radiology
One of the most structurally significant business developments in the diagnostic imaging AI space in January 2026 was not a device clearance but a commercial partnership: Bristol Myers Squibb partnered with Microsoft to deploy FDA-cleared radiology AI algorithms via Microsoft’s Precision Imaging Network for the early detection of lung cancer. The arrangement brings a major pharmaceutical company into the radiology AI ecosystem in an operational role — not as a technology developer but as a clinical outcome stakeholder willing to fund AI diagnostic deployment to support its own oncology therapeutic pipeline. This dynamic is likely to accelerate: as companion diagnostics become standard requirements for oncology drug approvals, pharmaceutical companies have a direct commercial interest in ensuring that the imaging AI infrastructure capable of identifying appropriate patients is widely deployed and clinically validated. The BMS-Microsoft partnership may prove to be an early example of a broader pharma-funded radiology AI deployment model.
The Radiologist Shortage That Is Forcing the AI Adoption Curve
The clinical urgency behind radiology AI’s adoption is not merely technological enthusiasm. It is the practical reality of a radiology workforce that has not kept pace with the dramatic growth in imaging volume driven by an ageing population, expanding cancer screening guidelines, and the proliferation of imaging across emergency and outpatient care. The Imaging Wire’s reporting on the FDA’s updated AI device list noted that healthcare providers are increasingly adopting AI tools specifically to address radiologist shortages and enhance operational efficiency. A statistic that has become widely cited across the specialty: AI cuts hospital system MRI wait times by more than 50% in deployments where AI-assisted scan acquisition and reconstruction reduce both scan time and radiologist review burden. iRhythm’s Zio patch, separately, achieves a 99% agreement rate between AI-guided and cardiologist-validated arrhythmia interpretations at a 99% patient compliance rate — a clinical validation benchmark that is shifting the conversation about AI in cardiac imaging from whether it is accurate enough to how best to integrate it into clinical workflows.
The 95-97% Problem: 510(k) Clearance and What It Does and Doesn’t Guarantee
The diagnostic imaging AI market’s rapid regulatory growth comes with an analytical caveat that Pinggy’s 2026 AI medical imaging analysis raised directly: approximately 95 to 97% of radiology AI clearances use the 510(k) pathway — demonstrating substantial equivalence to a predicate device — rather than the more rigorous De Novo or PMA routes. The 510(k) pathway does not require randomised clinical trial evidence of clinical superiority or patient outcome improvement; it requires demonstration of substantial equivalence to an existing cleared device. That matters when interpreting what an FDA clearance actually guarantees about a radiology AI tool’s real-world diagnostic performance. The field is aware of this limitation and is actively working to address it: the American College of Radiology’s AI-LAB registry, multi-centre clinical validation studies across multiple cleared tools, and the FDA’s own evolving guidance on AI/ML-based software as a medical device are all part of the effort to build the clinical evidence base that the clearance volume alone cannot fully substitute for.
What the Diagnostic Imaging Market Looks Like as AI Becomes the Default
Constancy Researchers’ assessment: diagnostic imaging in 2026 is the clearest example in all of medtech of a market in which artificial intelligence has moved from experimental pilot to operational infrastructure. GE HealthCare’s 130 radiology AI authorisations, Siemens’ six-system May clearance, Philips’ Spectral CT Verida, Aidoc’s 60 million cases per year across 2,000 hospitals, and the FDA’s 92 AI device clearances in a single quarter collectively describe a regulatory and commercial ecosystem that is generating, validating, and deploying clinical AI at a pace no other medical specialty approaches. The 510(k) pathway concern is real and worth tracking, but it has not slowed the deployment momentum. What defines competitive advantage in this market in 2026 is not whether a company has AI in its imaging systems — all of the major manufacturers do — but whether its AI is integrated deeply enough into the workflow, validated rigorously enough for the clinical environment, and supported by the enterprise software ecosystem that health system CIOs require before they commit to platform-level vendor relationships.
Looking to navigate geopolitical risks and identify new industry opportunities?
Connect with Constancy Researchers for data-driven insights, geopolitical intelligence, and strategic consulting tailored to your business needs.
