Advances in risk prediction models for Glaucoma: An updated narrative review

Abstract

Glaucoma is a leading cause of irreversible blindness and is characterized by optic nerve atrophy and progressive visual field loss. Risk prediction models are crucial for early screening and personalized treatment by identifying high-risk individuals for timely intervention. Recent advances in machine learning and artificial intelligence have improved prediction accuracy by integrating complex multivariable data. Models incorporating clinical factors, such as intraocular pressure, optic nerve head morphology, retinal nerve fiber layer thickness, and family history, as well as imaging and genomic markers, have demonstrated strong performance using algorithms, such as random forests and support vector machines. Importantly, emerging models enable stratified risk assessments for specific glaucoma subtypes, including primary open-angle glaucoma, primary angle-closure glaucoma, and secondary glaucoma, thereby supporting targeted screening and subtype-specific prevention strategies. This review summarizes recent progress in glaucoma risk prediction models and their applications in epidemiology, subtype risk evaluation, and blindness prevention, along with challenges in model generalizability, with the aim of advancing early detection and personalized care.