Peer-reviewed large-scale, independent head-to-head study of eight automated retinal image analysis systems (ARIAS) on over 200,000 consecutive patient visits from English NHS
Overview
This independent comparison study of multiple ARIAS was conducted on 202,886 consecutive patient cases from the North East London Diabetic Eye Screening Programme. Eight of 25 invited and potentially eligible CE-marked systems for diabetic retinopathy detection from retinal images agreed to participate. EyeArt system by Eyenuk participated in the study and is indicated with participation key B in the publication.
Methods
ARIAS results were compared against expert grading by three graders according to a standard national protocol. ARIAS performance overall and by subgroups of age, sex, ethnicity, and index of multiple deprivation (IMD) were assessed against the reference standard, defined as the final human grade in the worst eye for referable diabetic retinopathy (primary outcome).
Results
EyeArt system demonstrated 97.8-99.2% sensitivity for referable DR across people of different ethnicities and ages. It was shown to have the highest efficiency (specificity) among the three ARIAS with sensitivity above 90% (Figure 1 in publication).
Conclusion
ARIAS showed high sensitivity for medium-risk and high-risk diabetic retinopathy in a real-world screening service, with equitable performance across population subgroups. ARIAS could provide a cost-effective solution to deal with the rising burden of screening for diabetic retinopathy by safely triaging for human grading, substantially increasing grading capacity and rapid diabetic retinopathy detection.
Link to full study: https://www.thelancet.com/journals/landig/article/PIIS2589-7500(25)00096-2/fulltext