Performance of an artificial intelligence automated system for diabetic eye screening in a large English population
Authors' conclusion
The performance of a commercially available deep-learning artificial intelligence system for identifying diabetic retinopathy in an English national Diabetic Eye Screening Programme is presented. Using the pre-defined settings artificial intelligence performance was highest when identifying diabetic retinopathy which requires an action by the screening programme.
Study context
- Population
- Diabetic eye screening episodes in the English Diabetic Eye Screening Programmeimage sets from 10,000 episodes
- Participants
- 9,817 participants
- Comparison
- AI automated grading vs manual grading of the national Diabetic Eye Screening Programme (reference standard)
- Outcome
- Sensitivity and specificity of the AI system for detecting DR vs manual grade
Methodological detailsconsecutive screening episodes, commercially available deep-learning AI software
- Consecutive screening episodes
- Commercially available deep-learning AI software
Performance
As reported in the abstract
Detection of no DR vs any DR
| Test or model | Sensitivity | Specificity |
|---|---|---|
| Artificial intelligence grading system | 69.7% | 92.2% |
Artificial intelligence grading system
Sensitivity69.7%
Specificity92.2%
Further findings
Detection of no or mild DR vs significant or referrable DR
- Cases in which the AI grade missed significant DR
- 0
Parts of this Brief are withheld: the automatic checks could not confirm them against the abstract.
Original publication
Performance of an artificial intelligence automated system for diabetic eye screening in a large English population. Diabet Med. Published 2 March 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →