LibraryFamily 1 · Diagnostic Imaging & Pathology
Autonomous diabetic retinopathy screening in primary care
ValidatedActive deploymentRobust evidence
Formally validated with a published Assurance Pack.
Provides an autonomous screening decision for more-than-mild diabetic retinopathy from retinal fundus images, referring positive cases to ophthalmology.
Plain-language summary
AIChoose an audience and generate a tailored summary on demand. The AI uses only what is on this page.
Clinical context
- Clinical problem
- Detect a condition
- Point of care
- Diagnosis
- Nature of AI output
- A classification
- Clinical specialty
- Other
- Care setting
- Primary care
- Patient population
- Adults with type 1 or type 2 diabetes attending community diabetic eye screening.
- Intended use
- Provides an autonomous screening decision for more-than-mild diabetic retinopathy from retinal fundus images, referring positive cases to ophthalmology.
Technology
- AI technique
- Deep learning, Computer vision
- Input data
- Medical imaging
- Output type
- Diagnosis
- Autonomy level
- Autonomous (acts without a human)
- Model provenance
- Vendor proprietary
- Model version
- IDx-DR v2.3
- Built on a general-purpose model
- No
Deployment
- Status
- Active deployment
- Country
- Netherlands
- Deployment date
- 1 June 2023
- Sites
- 5
Regulatory & governance
- EU AI Act risk tier
- High-risk
- High-risk basis
- Annex I — medical device
- Medical device
- Yes
- EU MDR class
- Class IIb
- CE marking
- CE marked
- FDA status
- De Novo
- ISO 14971 risk class
- Medium
- GDPR processing basis
- Consent
- GDPR DPIA
- Completed
- Data identifiability
- Pseudonymised
- Explainability method
- Post-hoc
- Human oversight model
- Autonomous diagnostic output; clinical governance defines escalation, and any ungradable image is referred to a human grader.
NICE evidence standards
- ESF tier
- Tier C — treat / diagnose / calculate risk
- Evidence category
- Category 3
Performance summary
- Headline metric
- Sensitivity
- Value
- 0.87
- Subgroup performance assessed
- Yes
- Known bias signals
- Image quality and gradability vary with cataract prevalence; performance audited across age bands.
Similar deployments
AILooking for close matches in the Library…
Assurance Pack
Validation artefact produced by the AIH Lab review panel.
- Criteria version
- 2026.1
- Result
- PASSED
- Issued
- 1 May 2023
- Panel chair
- —
Evidence records
Studies and evaluations attached to this use case.
Prospective clinical trial
Specificity: 0.9
Population: Primary-care diabetic cohort, 900 patients
Prospective clinical trial
Sensitivity: 0.87
Population: Primary-care diabetic cohort, 900 patients
Contributors
- Deploying organisation
- [demo] Amsterdam UMC · Hospital / health system · Netherlands
- AI vendor
- [demo] Digital Diagnostics · United States
- Product name
- IDx-DR