Artificial Intelligence for Early Detection of Diabetic Retinopathy
Year: 2026
Type: conference-paper
Abstract: Objectives The aim of this systematic review is to evaluate the role of artificial intelligence in the early detection of diabetic retinopathy, with emphasis on diagnostic performance, retinal imaging modalities, clinical validation, and realworld implementation in screening programs. Methods A PRISMA-guided systematic review of 32 peer-reviewed studies published between 2021 and 2025 was conducted using PubMed-indexed literature. Included studies evaluated AI-based retinal imaging systems for early diabetic retinopathy detection, reporting diagnostic accuracy metrics and clinical validation outcomes. Results Diabetic retinopathy (DR) was identified as a major microvascular complication of diabetes associated with systemic vascular damage and increased cardiovascular and renal risk, with disease severity closely related to glycemic control [1-6]. Retinal imaging modalities, including fundus photography and optical coherence tomography, remain central to early DR detection, while AI-based systems demonstrated superior accuracy for identifying subtle early retinal lesions [7–12]. AI-driven diagnostic approaches achieved high performance, with reported accuracy up to 99.18%, sensitivity of 98.15%, and specificity reaching 100% for vision-threatening disease [13]. Meta-analyses reported pooled sensitivity of 87.7% and specificity of 90.6%, with real-world negative predictive values ≥96% and positive predictive values ≥90% [14–16]. Clinical validation studies demonstrated AUROC values ranging from 0.88 to 0.99 for referable DR detection, with performance comparable or superior to ophthalmologists [17,18]. Evidence from real-world studies and meta-analyses supports the clinical reliability and readiness of AI-based DR screening systems [19,20]. Conclusions Artificial intelligence represents an effective and clinically validated approach for early diabetic retinopathy screening. Integration of AI with retinal imaging improves early detection and screening efficiency, particularly in resource-limited settings, although challenges related to implementation, ethics, and clinical integration remain.
Author: Kaušyla, Kazimieras
Language: English
Topic: Retinal Imaging and Analysis
Subfield: Radiology, Nuclear Medicine and Imaging
Field: Medicine
Domain: Health Sciences
Open Access status: green
OpenAlex ID: W7160459711