Artificial Intelligence in Optometry and Ophthalmology: Current Applications, Challenges, and Future Perspectives

Background: Healthcare delivery has been drastically changed by artificial intelligence (AI), especially in image-based specialties like ophthalmology and optometry. Automated interpretation of retinal pictures, optical coherence tomography (OCT) scans, and visual field assessments with excellent diagnostic accuracy has been made possible by recent developments in machine learning (ML) and deep learning (DL). Objective:The purpose of this review is to provide an overview of the present uses of AI in ophthalmology and optometry, highlight breakthroughs unique to India, address ethical and legal issues, and investigate potential future paths for AIdriven eye care. Methods: For studies published between 2005 and 2025, a thorough evaluation of the literature was carried out using the databases PubMed, Scopus, Web of Science, and Google Scholar. "Artificial intelligence," "deep learning," "ophthalmology," "optometry," "diabetic retinopathy," "glaucoma," "OCT," and "teleophthalmology" were among the keywords. Included were studies assessing the use of AI in illness monitoring, screening, diagnosis, and predictive analytics. Results: In cases of diabetic retinopathy, glaucoma, age-related macular degeneration, keratoconus, and retinopathy of prematurity, AI-based systems showed outstanding diagnostic performance. Sensitivity and specificity scores above 90% were found in a number of investigations. Rural and underprivileged areas now have easier access to eye care services because to AI-assisted teleophthalmology platforms. AI integration into community eye care programs has been greatly aided by Indian organizations like LV Prasad Eye Institute and Aravind Eye Care System. Conclusion: In ophthalmology and optometry, artificial intelligence has significant promise for enhancing diagnostic accuracy, clinical effectiveness, and accessibility. Widespread adoption is still hampered by issues with data privacy, algorithmic bias, interpretability, infrastructure, and governmental permission. Explainable AI, multicenter validation, federated learning, and equitable deployment methodologies should be the main areas of future growth.