What question did this study set out to answer?

This research aims to identify retinal features differentiating between normoglycemia, diabetes mellitus, and prediabetes among Asian Indians using AI techniques.

January 25, 2026

Diagnostic Accuracy of Artificial Intelligence-Enabled Retinal Biomarkers for Detecting Type 2 Diabetes and Prediabetes Among Asian Indians DART Study

Key Points

This research aims to identify retinal features differentiating between normoglycemia, diabetes mellitus, and prediabetes among Asian Indians using AI techniques.
Analyzed 273 retinal images from 139 participants with varied glycemic statuses.
Extracted 226 vessel tortuosity features from retinal images using machine vision.
Selected top features using statistical methods for training and validation.
Evaluated AI model performance using precision-recall and receiver operating characteristic curves.
Assessed associations of oculomics scores with glycemic status through logistic regression.
Distinct retinal features identified for diabetes and prediabetes detection.
Vein-only model achieved high precision (AUPRC = 0.96) for diabetes detection.
For prediabetes, model shows moderate performance (AUC = 0.80).
Oculomic scores were significantly associated with diabetes but not prediabetes, suggesting differing impact.

Abstract

Purpose: To identify retinal microvascular features that distinguish normoglycemia from diabetes mellitus (DM)Study-1 and prediabetes (PreDM)Study-2 among Asian Indians using artificial intelligence (AI); evaluate their diagnostic accuracy; and examine independent associations of oculomics scores Oculomic Diabetes Score (ODiS) and Prediabetes Score (OPreS) with glycemic status. Methods: We analyzed 273 retinal images from 139 participants (19 = normoglycemia, 100 = DM, 20 = PreDM; mean age: 49.4 ± 12.9 yrs; male: 71.2%) with dataset randomly split (50:50) into training and test sets. We extracted 226 quantitative vessel tortuosity features separately for arteries and veins using machine vision-based approaches. Top six discriminating features were separately selected (using Wilcoxon Rank-Sum test and Linear Discriminant Analysis) on training sets and validated on independent blinded test sets. Model performances were evaluated using area under precision-recall curve (AUPRC) for DM and area under the receiver operating characteristic curve (AUC) for PreDM. Independent association of ODiS and OPreS (adjusting for age, sex, height, weight, blood pressure, serum cholesterol, serum creatinine) was assessed by multivariable logistic regression. Results: Specific oculomics-based retinal vascular features distinguished DM/PreDM from normoglycemia. Vein-only model achieved AUPRC = 0.96 (95% confidence interval 95% CI: 0.9–1.00) (sensitivity = 95%, specificity = 72.2%, precision = 95%) on S v1 for DM and an AUC = 0.80 (95% CI: 0.63–0.94) (sensitivity = 70%, specificity = 80%, precision = 82.35%) on S v2 for PreDM. Oculomics scores were independently associated with DM(ODiS) Adjusted Odds ratio (AdjOR): 2.00 (95% CI: 1.56–2.57, P < 0.001) but not with PreDM(OPreS) AdjOR: 1.32 (95% CI: 0.65–2.71, P = 0.48). Conclusions: In this proof-of-concept study, AI-informed retinal venous features on routine fundus images, with further prospective and multisite validation, could potentially serve as noninvasive DM detection using AI models among Asian Indians.

Bookmark

Cite This Study

Kaup et al. (Fri,) studied this question.

synapsesocial.com/papers/6975b38dfeba4585c2d6f0fe https://doi.org/https://doi.org/10.1177/15209156261417296

Bookmark