Primary care seeking suppliers for AI support for diabetic eye screening
A health service is seeking AI software for diabetic retinography in primary care, underscoring push to embed imaging AI in routine pathways.
A health service is seeking AI software for diabetic retinography in primary care, underscoring push to embed imaging AI in routine pathways.
Follow Tenderlake on LinkedIn for concise insights on public-sector tenders and emerging procurement signals.
A regional health service in northern Spain plans to bring artificial intelligence into diabetic eye screening in primary care. The move is set out in a prior information notice on AI software for diabetic retinography published in November 2025. It signals intent to supply and implement clinical decision-support software, based on AI, to assist diagnosis from retinography in the primary care setting.
The Servicio de Salud del Principado de Asturias has not released detailed specifications. As a prior information notice, the publication outlines the direction but not the full technical or contractual scope. The core aim is clear: introduce AI-enabled clinical diagnostic support for diabetic retinography within primary care workflows.
This is a pragmatic place to start. Retinography is already a standard part of diabetic eye screening. Applying AI at the point where images are captured and first assessed can support consistent grading and help route cases that need specialist attention. While the Asturian notice does not spell out functions, recent public projects elsewhere offer a guide to what health systems look for when trialling and deploying these tools.
In Catalonia, a market consultation in September 2022 set out a detailed ambition for AI in diabetic retinopathy across primary care. The Agència de Qualitat i Avaluació Sanitària de Catalunya described solutions to assess patient risk and determine ocular lesions and grading from clinical data and retinography images, with the goal of integration and interoperability in the public primary care system. The expected benefits were practical and patient‑centred: improved screening, higher coverage, personalised intervals between tests based on risk, fewer referrals to ophthalmology for borderline cases, earlier detection of serious injury with fewer associated treatments, and less travel to screening centres. That intent is captured in the September 2022 consultation notice.
Two years later, in October 2024, Catalonia moved from consultation to deployment with a contract notice to roll out AI solutions for diabetic retinopathy in primary care. That progression, from scoping to scale‑up, is documented in the October 2024 deployment notice. For Asturias, the parallel is not about copying a specification but about recognising a maturing pathway: define aims, test integration in primary care, and then scale.
Spain’s broader investment in primary care eye imaging provides a practical base for AI. In March 2023, the Valencian administration procured non‑mydriatic retinography equipment for primary care centres, as set out in this equipment notice. In November 2023, the Catalan Institute of Health acquired OCT retinographs for primary care, detailed in this notice. More recently, in June 2025, another Spanish health department sought an ophthalmic imaging system for retinopathy screening, as described here. The direction of travel is consistent: strengthen the imaging backbone in primary care, then add AI where it can support grading and triage.
Regions are not only buying devices. They are commissioning AI‑supported diagnostic services for ophthalmology. In May 2025, La Rioja initiated a service for robotic ophthalmological diagnostic tests using AI to support care at a university hospital, outlined in this notice. In July 2025, Cantabria sought AI‑enabled ophthalmological diagnostic testing services for a regional hospital, as set out here. And in October 2025, a hospital in Alicante planned around 7,000 AI‑based ophthalmological diagnostic tests per year, according to this service notice. These service models show how AI can be embedded as part of an end‑to‑end pathway rather than just as a feature on a device.
Elsewhere in Europe, buyers have started to lock in architectural choices that will frame AI adoption. In Norway, in January 2024, a national procurement body sought a cloud service for automated retinal image assessment for diabetic retinopathy screening, including components to manage information flow. The specification, captured in this January 2024 notice, points to a cloud‑first approach with attention to workflow integration.
Sweden’s Västra Götalandsregionen used an information request in January 2025 to explore AI solutions for screening diabetic retinopathy, as shown in this RFI. The focus there is market readiness and fit for purpose before committing to a specific route.
Some systems are opting for broader platforms that can host multiple AI use cases. In April 2022, a major German hospital network launched a procurement for a platform to manage AI solutions in clinical imaging, with initial use cases in fracture detection, brain haemorrhage and pulmonary embolism, and a plan to expand via a marketplace model. That approach is set out in this April 2022 notice. Spain has also looked at orchestration: in April 2024, a regional health service sought a device‑independent medical image capture tool and an AI orchestrator to support dermatology, documented here. For Asturias, the eventual specification will show whether it favours a focused algorithm for retinography, a broader platform, or an orchestrator to manage multiple tools.
Health services are extending AI into other imaging pathways. In March 2025, the regional health management in Castilla y León moved to acquire AI‑based software to support clinical diagnosis from mammography, as set out in this notice. In July 2025, Malta sought AI diagnosis systems for mammography alongside energy‑efficient computers and monitors for several hospitals, detailed here. And in Italy, in July 2025, a regional health trust sought AI software to study patients with suspected ischaemic stroke, as recorded in this procurement. The pattern is clear: once imaging AI beds in, it tends to spread across modalities.
The Asturian notice is intentionally concise. The next publication should clarify the intended deployment model and interoperability requirements. Recent notices suggest several points to look for:
For now, Asturias has put primary care diabetic retinography on the AI roadmap. With parallel Spanish and European projects moving from pilots to operations, the details that follow this prior information notice will show how quickly primary care can turn those ambitions into routine practice.
Follow Tenderlake on LinkedIn for concise insights on public-sector tenders and emerging procurement signals.