An Automated Multimodal Glaucoma Detection Framework Using ViT and a Stacking-Based Ensemble

2026-07-02

AI Summary

This study addresses the inefficiency and reliance on expert interpretation in early glaucoma screening by proposing a multimodal automatic detection framework that integrates fundus images with clinical data. The method leverages a Vision Transformer (ViT) to extract visual features and combines them with traditional machine learning models within a stacking ensemble architecture, enabling both sample-level and patient-level diagnosis. Evaluated on the PAPILA dataset, the model achieves 97.47% accuracy and 97.50% F1-score at the sample level, and 98.97% accuracy and F1-score at the patient level. By innovatively integrating ViT with a stacking ensemble strategy and deploying the system as an end-to-end web platform, this work significantly enhances the feasibility and precision of large-scale glaucoma screening.

Abstract

Glaucoma is a progressive eye disease that can lead to irreversible vision loss if not detected at an early stage. Conventional diagnostic procedures are often time-consuming and rely heavily on expert interpretation, limiting their scalability for large-scale screening. In this study, glaucoma detection is investigated under two evaluation settings: sample-wise, where individual samples are analyzed independently, and patient-wise, where data from each patient are aggregated for final prediction. An automated multimodal framework is proposed that integrates fundus images with clinical data. Under the sample-wise setting, detection is performed using fundus images and clinical features individually, as well as through their multimodal combination. Under the patient-wise setting, predictions are obtained by aggregating multiple fundus image representations with corresponding clinical information for each patient. Deep visual features are extracted using a Vision Transformer (ViT) architecture and classified using classical machine-learning models, with a stacking-based ensemble of the three best-performing classifiers employed to optimize performance. Experiments conducted on the publicly available PAPILA dataset demonstrate strong diagnostic performance, achieving 97.47% accuracy and a 97.50% F1-score for sample-wise multimodal classification, and 98.97% accuracy and F1-score for subject-wise detection. The proposed framework is further deployed as an end-to-end web-based platform to support automated glaucoma screening and clinical decision support.

Problem

• glaucoma detection
• automated diagnosis
• multimodal framework
• large-scale screening
• early detection

Innovation

• Vision Transformer
• multimodal fusion
• stacking ensemble
• glaucoma detection
• patient-wise aggregation