Search

Background/aims Human grading of digital images from diabetic retinopathy (DR) screening programmes represents a significant challenge, due to the increasing prevalence of diabetes. We evaluate the performance of an automated artificial intelligence (AI) algorithm to triage retinal images from the English Diabetic Eye Screening Programme (DESP) into test-positive/technical failure versus test-negative, using human grading following a standard national protocol as the reference standard. Methods Retinal images from 30 405 consecutive screening episodes from three English DESPs were manually graded following a standard national protocol and by an automated process with machine learning enabled software, EyeArt v2.1. Screening performance (sensitivity, specificity) and diagnostic accuracy (95% CIs) were determined using human grades as the reference standard. Results Sensitivity (95% CIs) of EyeArt was 95.7% (94.8% to 96.5%) for referable retinopathy (human graded ungradable, referable maculopathy, moderate-tosevere non-proliferative or proliferative). This comprises sensitivities of 98.3% (97.3% to 98.9%) for mild-tomoderate non-proliferative retinopathy with referable maculopathy, 100% (98.7%,100%) for moderate-tosevere non-proliferative retinopathy and 100% (97.9%,100%) for proliferative disease. EyeArt agreed with the human grade of no retinopathy (specificity) in 68% (67% to 69%), with a specificity of 54.0% (53.4% to 54.5%) when combined with non-referable retinopathy. Conclusion The algorithm demonstrated safe levels of sensitivity for high-risk retinopathy in a real-world screening service, with specificity that could halve the workload for human graders. AI machine learning and deep learning algorithms such as this can provide clinically equivalent, rapid detection of retinopathy, particularly in settings where a trained workforce is unavailable or where large-scale and rapid results are needed. [ABSTRACT FROM AUTHOR]

Background: Photographic diabetic retinopathy screening requires labour-intensive grading of retinal images by humans. Automated retinal image analysis software (ARIAS) could provide an alternative to human grading. We compare the performance of an ARIAS using true-colour, wide-field confocal scanning images and standard fundus images in the English National Diabetic Eye Screening Programme (NDESP) against human grading. Methods: Cross-sectional study with consecutive recruitment of patients attending annual diabetic eye screening. Imaging with mydriasis was performed (two-field protocol) with the EIDON platform (CenterVue, Padua, Italy) and standard NDESP cameras. Human grading was carried out according to NDESP protocol. Images were processed by EyeArt V.2.1.0 (Eyenuk Inc, Woodland Hills, California). The reference standard for analysis was the human grade of standard NDESP images. Results: We included 1257 patients. Sensitivity estimates for retinopathy grades were: EIDON images; 92.27% (95% CI: 88.43% to 94.69%) for any retinopathy, 99% (95% CI: 95.35% to 100%) for vision-threatening retinopathy and 100% (95% CI: 61% to 100%) for proliferative retinopathy. For NDESP images: 92.26% (95% CI: 88.37% to 94.69%) for any retinopathy, 100% (95% CI: 99.53% to 100%) for vision-threatening retinopathy and 100% (95% CI: 61% to 100%) for proliferative retinopathy. One case of vision-threatening retinopathy (R1M1) was missed by the EyeArt when analysing the EIDON images, but identified by the human graders. The EyeArt identified all cases of vision-threatening retinopathy in the standard images. Conclusion: EyeArt identified diabetic retinopathy in EIDON images with similar sensitivity to standard images in a large-scale screening programme, exceeding the sensitivity threshold recommended for a screening test. Further work to optimise the identification of 'no retinopathy' and to understand the differential lesion detection in the two imaging systems would enhance the use of these two innovative technologies in a diabetic retinopathy screening setting. [ABSTRACT FROM AUTHOR]