Effectiveness of AI-Based Tools in Detecting Diabetic Retinopathy in Low- and Middle-Income Countries: A Systematic Review of Diagnostic Performance and Implementation Feasibility.

Diabetic retinopathy (DR) is a leading cause of preventable blindness worldwide, with a disproportionate impact in low- and middle-income countries (LMICs). Artificial intelligence (AI) offers a potential means to address workforce and infrastructure gaps that limit access to DR screening in these settings, but evidence on its performance and feasibility remains scattered. A systematic review of studies published between January 2015 and June 2025 was conducted using six databases. Eligible studies evaluated AI, machine learning, or deep learning applied to retinal imaging for DR detection and reported quantitative diagnostic or implementation outcomes, while studies limited to high-income countries or non-original research were excluded. Only a small number of eligible studies were identified. Across these, AI-based tools generally showed high diagnostic accuracy and were feasible to implement in resource-limited environments. Early evidence suggested potential benefits, such as reduced screening costs, decreased clinician workload, and improved patient follow-up, though reporting on infrastructure needs, regulatory considerations, and long-term sustainability was limited. Overall, AI-based tools show promise for scaling DR screening in LMICs, with encouraging indications of good accuracy and operational efficiency, but further large-scale and implementation-focused research is required to guide their integration into health systems.