Can smartphone images be used to screen for diabetic retinopathy in low-resource settings?

Smartphone funduscopy uses a lens attachment and a smartphone camera to take pictures of the back of the eye (the retina). These images can be graded by trained readers or by artificial intelligence (AI) software to check for signs of diabetic retinopathy. In low-resource settings where access to eye specialists and expensive imaging equipment is limited, smartphone-based screening offers a portable, lower-cost alternative. Research shows that these systems can detect diabetic retinopathy with good accuracy, making them a promising tool for expanding screening coverage.

What the research says

A 2024 review of smartphone funduscopy systems found that they are particularly useful for automating and mobilizing retinopathy screening in remote and rural areas, helping to reduce patient travel, appointment backlogs, and the workload on eye specialists ⁹. Another review from 2015 noted that several smartphone adapters and techniques have been developed, and some systems have shown good agreement with standard retinal cameras, but more large-scale testing in real screening programs is needed before they can be widely adopted ¹⁰.

A 2025 meta-analysis of 45 studies found that AI-based screening tools, including those using smartphone images, were strongly associated with detecting diabetic retinopathy, with a pooled odds ratio of 5.79 (95% CI: 5.22–6.42) ⁵. A separate study developed a deep learning model specifically for low-resource settings, training it on both high-quality clinical images and lower-quality smartphone images. The best model achieved an AUC of 87%, sensitivity of 89%, and specificity of 72% on smartphone images ⁶.

Other research has explored simpler risk prediction tools that use only routinely collected health data (like diabetes duration and HbA1c) to identify patients at high risk for diabetic retinopathy, which could complement smartphone imaging in primary care ⁷. However, a cross-sectional study in Somaliland found that about 51% of diabetes patients had diabetic retinopathy, highlighting the high burden and need for accessible screening in low-resource settings ⁸.

What to ask your doctor

• Is smartphone-based retinal imaging available at our clinic or through a telemedicine program?
• How accurate are the smartphone screening tools compared to a standard eye exam for diabetic retinopathy?
• If I have diabetes, how often should I be screened for diabetic retinopathy, and can smartphone imaging replace a dilated eye exam?
• What should I do if a smartphone screening suggests I might have diabetic retinopathy?
• Are there any ongoing studies or programs in our area that use smartphone screening for diabetic retinopathy?