Search

Objective: To identify age-related vascular changes in the optic discs of patients with diabetes with and without signs of glaucoma., Methods and Analysis: A total of 2153 eyes of 1797 patients with diabetes without significant retinopathy were monitored with 10 Topcon-NW400 images obtained over 10.27±1.58 years. 571 non-diabetics eyes were selected as controls. Laguna ONhE uses convolutional neural networks to identify optic disc edges, vessels, cup and rim, and provides a glaucoma assessment index-Globin Distribution Function (GDF)., Results: In the first image, vessel pixels accounted for 33.88% of the disc area (SD=3.72) in non-glaucoma (DN) and 31.35% (SD=4.05; p<0.0001) in glaucoma cases (DG). This number of pixels was reduced by -0.55% each year (SD=0.77) in the DN and -0.76% (SD=0.86; p=0.0014) in the DG. In the first image, 76.55% of the disc pixels (SD=11.13) belonged to the rim in the DN and 62.05% (SD=11.00; p=0.0014) in the DG, decreasing annually by -0.33% (SD=0.99) in the DN and -0.68% (SD=1.08; p<0.00001) in the DG groups. All rim sectors were reduced over time in the DG group, particularly superotemporal (41°-80°) and inferotemporal (271°-310°). The reduction was smaller in DN, presenting as progressive thickening of the temporal sector (311°-40°). No changes in age were observed in healthy controls., Conclusion: Patients with diabetes show progressive reduction of vessels and neuroretinal rim at the optic disc, which is more intense in association with glaucoma. In the absence of glaucoma, the temporal sector of the diabetic rim was not reduced but thickened, displacing the cup nasally., Competing Interests: Competing interests: DG-H, MG-H and MGdlR participated in the patent rights of the Laguna ONhE method and its partners. The rest of the authors have no conflicts of interest to declare., (© Author(s) (or their employer(s)) 2022. Re-use permitted under CC BY-NC. No commercial re-use. See rights and permissions. Published by BMJ.)

Aims: To assess the effectiveness of a non-mydriatic digital camera (45 degrees -30 degrees photographs) compared with the reference method for screening diabetic retinopathy., Methods: Type 1 and 2 diabetic patients (n = 773; 1546 eyes) underwent screening for diabetic retinopathy in a prospective observational study. Hospital-based non-mydriatic digital retinal imaging by a consultant specialist in retinal diseases was compared with slit-lamp biomicroscopy and indirect ophthalmoscopy through dilated pupils, as a gold standard, previously performed in a community health centre by another consultant specialist in retinal diseases. The main outcome measures were sensitivity and specificity of screening methods and prevalence of diabetic retinopathy., Results: The prevalence of any form of diabetic retinopathy was 42.4% (n = 328); the prevalence of sight-threatening including macular oedema and proliferative retinopathy was 9.6% (n = 74). Sensitivity of detection of any diabetic retinopathy by digital imaging was 92% (95% confidence interval 90, 94). Specificity of detection of any diabetic retinopathy was 96% (95, 98). The predictive value of the negative tests was 94% and of a positive test 95%. For sight-threatening retinopathy digital imaging had a sensitivity of 100%., Conclusions: A high sensitivity and specificity are essential for an effective screening programme. These results confirm digital retinal imaging with a non-mydriatic camera as an effective option in community-based screening programmes for diabetic retinopathy.