A recent study published in the journal The Lancet reported that an artificial intelligence (AI)–based eye screening tool successfully detected glaucoma during a village eye camp, highlighting the potential of AI to improve early diagnosis of eye diseases in resource-limited settings.
Glaucoma is a chronic eye disease that damages the optic nerve and is one of the leading causes of irreversible blindness worldwide. Because the disease often develops slowly and without noticeable symptoms in its early stages, many individuals remain undiagnosed until vision loss has already occurred.
Traditional diagnosis relies on several clinical tests, including optic nerve examination, visual field testing, and imaging of the retina. However, these methods require trained specialists and may not always be accessible in large-scale screening settings. This has led researchers to explore artificial intelligence (AI) as a tool to assist in earlier and more accessible detection of glaucoma.
A recent study published in The Lancet Regional Health – Western Pacific investigated the feasibility and diagnostic performance of integrating AI-based glaucoma screening into primary eye care systems. The research assessed how automated analysis of retinal images could help identify individuals who may require further evaluation by ophthalmologists.
According to the study, AI algorithms can analyze retinal photographs and detect structural changes associated with glaucoma, including abnormalities in the optic nerve head. The researchers also examined whether AI-assisted screening could improve efficiency and potentially support earlier treatment, which is important for preventing avoidable vision loss.
The study focused on evaluating diagnostic accuracy and the potential role of AI in population-based screening programs.
AI systems designed for glaucoma detection typically rely on deep learning models trained using thousands of retinal images. These algorithms analyze patterns within images captured during eye examinations, such as fundus photography or optical coherence tomography (OCT).
Key features assessed by AI systems include:
Changes in the optic nerve head
Enlargement of the optic cup (cup-to-disc ratio)
Retinal nerve fiber layer thinning
Structural abnormalities in the retina
By identifying these features, AI models can classify images as normal or suspicious for glaucoma, helping prioritize patients for specialist evaluation.
Multiple studies have explored the use of AI in glaucoma detection. Research using deep learning models applied to retinal images has reported high diagnostic accuracy in identifying glaucomatous changes.
In some clinical evaluations, AI-based systems demonstrated sensitivity and specificity levels exceeding 90% when detecting glaucomatous optic nerve damage.
A meta-analysis examining AI models applied to fundus photographs and OCT imaging also found strong diagnostic performance across different datasets.
Other studies have shown that combining AI algorithms with clinical imaging techniques may support ophthalmologists in identifying glaucoma more efficiently.
AI-powered screening tools are increasingly being studied for use in community health programs and teleophthalmology services. Automated systems can analyze retinal images remotely, potentially enabling screening in regions where ophthalmology specialists are limited.
For example, AI algorithms can process images captured by portable fundus cameras or even smartphone-based devices, helping identify individuals who require referral for further examination.
This approach could support large-scale screening initiatives and may help detect glaucoma earlier in populations where routine eye examinations are less common.
Reference
Y. Zhang et al., “Artificial Intelligence–Based Glaucoma Screening in Primary Eye Care: A Prospective Diagnostic Study,” The Lancet Regional Health – Western Pacific, 2026, https://doi.org/10.1016/S3050-5143(26)00002-6.
(Rh)
