Detecting Glaucoma Across Multi-ethnic Myopic and Non-Myopic Populations Using an Uncertainty-Aware Vision Transformer: A Multicentre Model Development and Validation Study

2026-09-24

AI Summary

This study addresses the performance degradation of AI-based glaucoma detection in multi-ethnic and highly myopic populations caused by data distribution shifts. We develop an uncertainty-aware model based on the ViT-B/16 architecture, trained on 56,000 fundus images and validated across 16 independent global datasets. The core contributions include the first unified modeling framework encompassing multi-ethnic cohorts as well as both myopic and non-myopic populations, alongside the integration of predictive uncertainty estimation to enhance clinical trustworthiness. Experimental results demonstrate an internal AUROC of 98.7% and external generalization ranging from 86.4% to 99.6%. Notably, the model significantly outperforms general ophthalmologists in diagnostic accuracy within myopic scenarios, exhibiting exceptional cross-population robustness.

Abstract

Background: Artificial intelligence (AI)-based glaucoma detection from colour fundus photographs (CFP) offers scalable screening, but performance may decline on external datasets because of differences in ground-truth definitions, populations, and coexisting conditions such as high myopia (HM). We developed and validated a Vision Transformer-based deep learning (DL) model for glaucoma detection across multi-ethnic cohorts with and without HM. Methods: A ViT-B/16 model with predictive uncertainty estimation was developed using 56,483 CFPs (57.1% with myopia; 14.4% with HM). Glaucoma labels were standardised using clinical, imaging, and perimetry data. The model was validated on 16 independent datasets across three continents, including four datasets with explicit HM labels. Findings: Internal AUROC was 98.7% (95% CI 98.2-99.1%), with sensitivity 94.5% and specificity 97.3%. Across 16 external datasets from eight countries, AUROCs ranged from 86.4% to 99.6%. In HM eyes, internal AUROC was 97.8% (95% CI 96.1-99.2%), with sensitivity 94.8% and specificity 93.7%. External HM AUROCs were 86.5% in the Beijing Eye Study and 93.3%, 91.8%, and 85.5% in hospital-based datasets from Taiwan, Thailand, and South Korea. In an exploratory HM clinical evaluation, the model had higher CFP-only diagnostic accuracy than ophthalmologists and trained graders (92.0% vs 70.0%; p=0.008) and performed comparably to glaucoma specialists using full clinical information. Interpretation: The model showed robust glaucoma detection across myopic and non-myopic multi-ethnic populations and may support AI-assisted screening in settings with high HM prevalence.

Problem

• Glaucoma detection
• High myopia
• Multi-ethnic populations
• External validation
• Color fundus photographs

Innovation

• Vision Transformer
• Uncertainty Estimation
• Glaucoma Detection
• High Myopia
• Multicentre Validation