Deep Learning and Machine Learning for Glaucoma Screening in Retinal Fundus Images: A Scoping Review Protocol and Methodological Taxonomy

This project registers the formal PRISMA-ScR protocol for a comprehensive Scoping Review mapping Machine Learning (ML) and Deep Learning (DL) architectures applied to automated glaucoma screening, staging, and anatomical biomarker extraction in retinal fundus photographs. Combining an automated Scopus query focused on the 2023–2026 technological frontier with backward citation snowballing for foundational baselines, the review synthesizes over 130 peer-reviewed studies across four methodological axes: (1) multi-comorbidity ocular classification, (2) multimodal data fusion, (3) secondary prediction via anatomical segmentation/regression, and (4) primary end-to-end prediction.