In this Article, we present a real-world evaluation of licensed Automated retinal image analysis systems (ARIAS) in a large, ethnically diverse screening programme.
Eight of 25 invited and potentially eligible CE-marked systems for diabetic retinopathy detection from retinal images agreed to participate.
From 202,886 screening encounters at the North East London Diabetic Eye Screening Programme (between Jan 1, 2021, and Dec 31, 2022) we curated a database of 1.2 million images and sociodemographic and grading data.
Sensitivity across vendors ranged from 83.7% to 98.7% for referable diabetic retinopathy, from 96.7% to 99.8% for moderate-to-severe non proliferative diabetic retinopathy, and from 95.8% to 99.5% for proliferative diabetic retinopathy.
Sensitivity was largely consistent for moderate-to-severe non-proliferative and proliferative diabetic retinopathy by subgroups of age, sex, ethnicity, and IMD for all ARIAS.
False positive rates for no observable diabetic retinopathy ranged from 4.3% to 61.4% and within vendors varied by 0.5 to 44 percentage points across population subgroups.
ARIAS showed high sensitivity for medium-risk and high-risk diabetic retinopathy in a real-world screening service, with equitable performance across population subgroups.
【MORE】