What happened
Region Skåne's Method and Priority Council asked whether the region should use AI to grade fundus photos of people with diabetes and no known retinopathy. It drew on a December 2025 review by HTA South, the assessment center for southern Sweden's health care region. On 22 April 2026, the council recommended that the region may use the software to pick out people who can keep their ordinary screening interval. It judged that the software can replace manual review with at least equal precision. That finding rests on four studies, which HTA South rated as low in certainty. Two used DAIRET by Retmarker, one used Aireen and one used an in-house model. The council expected about 1,400 hours a year to be freed for other eye care across four regions. It expected little or no change for the people screened. It could draw no conclusions on vision loss, blindness, recalls or cost-effectiveness. The council rated the evidence as low to moderate in quality and asked for structured follow-up of any adoption.
Sources
-
Assessment report, by authors with no tie to the system.
Region Skåne, Metod- och prioriteringsrådet. Yttrande: AI-utvecklad automatiserad bildanalys av ögonbottenfotografier vid screening för diabetesretinopati. Dnr 2026-O000003, 22 April 2026. Metod- och prioriteringsrådets ställningstagande.
Region Skånes Metod- och prioriteringsråd rekommenderar att AI-utvecklad automatiserad bildanalys av ögonbottenfotografier kan användas vid screening av personer med diabetes utan tidigare känd diabetesretinopati för att identifiera personer som skall fortsätta kontroller med ordinarie screeningintervall.
-
Assessment report, by authors with no tie to the system.
HTA syd (HTA South), Region Skåne. AI-baserad bildanalys vid screening för diabetesretinopati: en systematisk översikt och utvärdering av medicinska, hälsoekonomiska, etiska och organisatoriska aspekter. Rapport 2025:4, Lund, 17 December 2025. English abstract, Key Message.
The health economic analysis assessed a scenario where replacing manual grading with AI could save 1,400 staff hours annually in the Southern Healthcare Region.
Details
AI system
- Workflow
-
The software graded fundus photographs that a nurse or ophthalmologist would otherwise review. The studies compared the software's grading with grading by at least one person with relevant specialist competence.
Follow-up
- Conditions for replication
-
Each clinic integrates the software with its cameras and its medical records system. The region follows up any adoption to track recalls, treatment and vision loss over time.
Entry details
- Sector
-
Healthcare.
- Purpose
-
Retmarker, Aireen and other makers built AI software to grade fundus photographs for diabetic retinopathy and to sort out who needs referral.
30 more questions not yet checked
model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, beneficiaries, reach, distribution and access, duration, every measurement question, every money question, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, other identifiers, category mappings.
Cite this record
Peter H. Vartanian. (2026). Skåne backs AI fundus grading to free hours in diabetic eye screening (AIBD-1124). Artificial Intelligence Benefits Database, edition of 1 November 2026. https://petervartanian.xyz/aibd/records/AIBD-1124.html (archived at https://doi.org/10.7910/DVN/AADSII, version 7.0)