Diagnostic Accuracy of Artificial Intelligence-Enabled Retinal Biomarkers for Detecting Type 2 Diabetes and Prediabetes Among Asian Indians [DART Study]
Authors' conclusion
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.
Study context
- Population
- Asian Indians; mean age 49.4 ± 12.9 yrs; male 71.2%; 273 retinal images
- Participants
- 139 participants273 retinal images
- Comparison
- Artificial intelligence (AI)-derived retinal vessel tortuosity features for diabetes mellitus (DM) and PreDM vs normoglycemia
- Normoglycemian = 19reference
- DMn = 100
- PreDMn = 20
- Outcome
- Diagnostic accuracy of vein-only model
Methodological detailsproof-of-concept, training and independent blinded test sets (random 50:50 split)
- Proof-of-concept
- Training and independent blinded test sets (random 50:50 split)
Performance
As reported in the abstract
| Test or model | Value | Sensitivity | Specificity | Precision |
|---|---|---|---|---|
| DM | Area under precision-recall curve (AUPRC) 0.96 (95% CI 0.9–1.00) | 95% | 72.2% | 95% |
| PreDM | AUC 0.80 (95% CI 0.63–0.94) | 70% | 80% | 82.35% |
Value
DMAUPRC 0.96 (95% CI 0.9–1.00)
PreDMAUC 0.80 (95% CI 0.63–0.94)
Sensitivity
DM95%
PreDM70%
Specificity
DM72.2%
PreDM80%
Precision
DM95%
PreDM82.35%
Further findings
Independent association of oculomics scores with glycemic status (multivariable logistic regression)
| Adjusted OR (95% CI) | P value | ||
|---|---|---|---|
| ODiS (DM) | 2.00 (1.56–2.57) | P < 0.001 | |
| OPreS (PreDM) | 1.32 (0.65–2.71) | P = 0.48 | |
| 0.20.5125Adjusted odds ratio (log scale) |
- Independent association of oculomics scores with glycemic status (multivariable logistic regression): adjusted for age, sex, height, weight, blood pressure, serum cholesterol, serum creatinine
Original publication
Diagnostic Accuracy of Artificial Intelligence-Enabled Retinal Biomarkers for Detecting Type 2 Diabetes and Prediabetes Among Asian Indians [DART Study]. Diabetes Technol Ther. Published 23 January 2026. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →