AI for Glaucoma Detection Using OCT
Wednesday, February 11 2026 | 08 h 36 min | Optik Magazine, Vision Science
Based on: Ophthalmology Science, 2025 – Deep Learning Study
At a Glance
- Population: OCT and fundus image datasets, thousands of eyes
- Technology: Deep learning models applied to OCT vs. disc photos
- Key outcome: Accuracy in detecting visual field–defined glaucoma
Summary
Australian researchers tested deep learning models trained on OCT versus traditional disc photographs to detect visual field–defined glaucoma. OCT-based AI models significantly outperformed disc-photo models, reinforcing the value of structural data for early disease detection.
While not ready for clinical rollout, the findings point to a future where OCT integrated with AI could assist with triage, referrals, and missed diagnoses. For eye care professionals, the study highlights OCT’s central role in glaucoma management, and the potential of AI to support, not replace, clinical expertise.
Practice Considerations
- OCT continues to prove its value as the structural foundation in glaucoma care.
- AI-assisted screening tools may soon enhance triage, referrals, and early detection.
- Models still require validation across devices and patient populations before adoption.
- Human expertise remains central—AI serves as augmentation, not replacement.
