Abstract

Purpose
To compare the performance of a vision transformer-based foundation model (RETFound) and a supervised convolutional neural network (VGG-19) for detecting referable glaucoma from fundus photographs.
Design
An evaluation of diagnostic technology.
Participants
Six thousand one hundred sixteen participants from the Los Angeles County Department of Health Services Teleretinal Screening Program.
Methods
Fundus photographs were labeled for referable glaucoma (cup-to-disc ratio ≥0.6) by certified optometrists. Four deep learning models were trained on cropped and uncropped images (training N=8996; validation N=3002) using 2 architectures: RETFound, a vision transformer with self-supervised pretraining on fundus photographs, and VGG-19. Models were evaluated on a held-out test set (N = 1000) labeled by glaucoma specialists and an external test set (N = 300) from University of Southern …

Date

2026

Authors

Kyle Bolo, Tran Huy Nguyen, Sreenidhi Iyengar, Zhiwei Li, Van Nguyen, Brandon J Wong, Jiun L Do, Jose-Luis Ambite, Carl Kesselman, Lauren P Daskivich, Benjamin Y Xu

Journal

Ophthalmology Science

Volume

6

Issue

2

Publisher

Elsevier