Research Information System
Journal of Ophthalmology, vol.2025, no.1, 2025 (SCI-Expanded, Scopus)
Purpose: To propose a conceptual framework for an AI-driven personalized glaucoma progression prediction system that integrates diverse data sources to enhance clinical decision-making and improve patient outcomes. This framework is specifically designed to address healthcare disparities in glaucoma care through scalable AI technology that can function across diverse resource settings, from tertiary care centers to remote clinics. The system aims to democratize access to expert-level glaucoma care while addressing challenges of bias, equity, and accessibility. Methods: The paper outlines a comprehensive framework consisting of four main components: (1) data integration and preprocessing, (2) AI model architecture and training, (3) personalized prediction generation, and (4) a clinical decision support interface. The framework leverages multimodal neural networks to analyze structural imaging data, functional test results, clinical measurements, and patient demographics. Results: The proposed framework addresses current limitations in glaucoma progression prediction by capturing complex interactions between various risk factors. Potential benefits include early detection of rapid progressors, optimized treatment strategies, improved patient counseling, and support for clinical research. Implementation challenges such as data quality, model interpretability, workflow integration, regulatory approval, and ethical considerations are discussed along with strategies to address them. Conclusions: The AI-driven framework for glaucoma progression prediction represents a significant advancement in personalized glaucoma management. While challenges remain, the potential benefits in terms of preserved vision, improved quality of life, and more efficient healthcare delivery are substantial. Future research directions include incorporating genetic data, advanced imaging modalities, and federated learning techniques to further enhance the system’s capabilities and impact.