Artificial intelligence for teleophthalmology-based diabetic retinopathy screening in a national programme: an economic analysis modelling study.

BACKGROUND: Deep learning is a novel machine learning technique that has been shown to be as effective as human graders in detecting diabetic retinopathy from fundus photographs. We used a cost-minimisation analysis to evaluate the potential savings of two deep learning approaches as compared with the current human assessment: a semi-automated deep learning model as a triage filter before secondary human assessment; and a fully automated deep learning model without human assessment. METHODS: In this economic analysis modelling study, using 39 006 consecutive patients with diabetes in a national diabetic retinopathy screening programme in Singapore in 2015, we used a decision tree model and TreeAge Pro to compare the actual cost of screening this cohort with human graders against the simulated cost for semi-automated and fully automated screening models. Model parameters included diabetic retinopathy prevalence rates, diabetic retinopathy screening costs under each screening model, cost of medical consultation, and diagnostic performance (ie, sensitivity and specificity). The primary outcome was total cost for each screening model. Deterministic sensitivity analyses were done to gauge the sensitivity of the results to key model assumptions. FINDINGS: From the health system perspective, the semi-automated screening model was the least expensive of the three models, at US$62 per patient per year. The fully automated model was $66 per patient per year, and the human assessment model was $77 per patient per year. The savings to the Singapore health system associated with switching to the semi-automated model are estimated to be $489 000, which is roughly 20% of the current annual screening cost. By 2050, Singapore is projected to have 1 million people with diabetes; at this time, the estimated annual savings would be $15 million. INTERPRETATION: This study provides a strong economic rationale for using deep learning systems as an assistive tool to screen for diabetic retinopathy. FUNDING: Ministry of Health, Singapore.

Quantitative assessment of early diabetic retinopathy using fractal analysis.

OBJECTIVE: Fractal analysis can quantify the geometric complexity of the retinal vascular branching pattern and may therefore offer a new method to quantify early diabetic microvascular damage. In this study, we examined the relationship between retinal fractal dimension and retinopathy in young individuals with type 1 diabetes. RESEARCH DESIGN AND METHODS: We conducted a cross-sectional study of 729 patients with type 1 diabetes (aged 12-20 years) who had seven-field stereoscopic retinal photographs taken of both eyes. From these photographs, retinopathy was graded according to the modified Airlie House classification, and fractal dimension was quantified using a computer-based program following a standardized protocol. RESULTS: In this study, 137 patients (18.8%) had diabetic retinopathy signs; of these, 105 had mild retinopathy. Median (interquartile range) retinal fractal dimension was 1.46214 (1.45023-1.47217). After adjustment for age, sex, diabetes duration, A1C, blood pressure, and total cholesterol, increasing retinal vascular fractal dimension was significantly associated with increasing odds of retinopathy (odds ratio 3.92 [95% CI 2.02-7.61] for fourth versus first quartile of fractal dimension). In multivariate analysis, each 0.01 increase in retinal vascular fractal dimension was associated with a nearly 40% increased odds of retinopathy (1.37 [1.21-1.56]). This association remained after additional adjustment for retinal vascular caliber. CONCLUSIONS: Greater retinal fractal dimension, representing increased geometric complexity of the retinal vasculature, is independently associated with early diabetic retinopathy signs in type 1 diabetes. Fractal analysis of fundus photographs may allow quantitative measurement of early diabetic microvascular damage.