MaxGlaViT: A novel lightweight vision transformer-based approach for early diagnosis of glaucoma stages from fundus images

2025-02-24

AI Summary

This study addresses the clinical challenge of early-stage glaucoma—characterized by asymptomatic progression and low diagnostic visibility—by proposing a lightweight, multi-stage automated grading framework. Methodologically, we introduce an enhanced MaxViT architecture: replacing the conventional convolutional stem with an ECA-enhanced stem, substituting MBConv blocks with ConvNeXtV2 and InceptionNeXt modules, and integrating Multi-Axis ViT with hybrid attention mechanisms (ECA, CBAM, and SE) to strengthen fine-grained disease-stage discrimination. Evaluated on the multi-center HDV1 fundus image dataset, our model achieves 92.03% accuracy, 92.33% precision, 92.03% recall, 92.13% F1-score, and 87.12% Cohen’s kappa—outperforming all 80 baseline models. The proposed framework establishes a novel, efficient, and deployable vision transformer paradigm for large-scale glaucoma screening.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Web Mining and Content Analysis: Large pretrained models with web data
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Glaucoma is a prevalent eye disease that progresses silently without symptoms. If not detected and treated early, it can cause permanent vision loss. Computer-assisted diagnosis systems play a crucial role in timely and efficient identification. This study introduces MaxGlaViT, a lightweight model based on the restructured Multi-Axis Vision Transformer (MaxViT) for early glaucoma detection. First, MaxViT was scaled to optimize block and channel numbers, resulting in a lighter architecture. Second, the stem was enhanced by adding attention mechanisms (CBAM, ECA, SE) after convolution layers to improve feature learning. Third, MBConv structures in MaxViT blocks were replaced by advanced DL blocks (ConvNeXt, ConvNeXtV2, InceptionNeXt). The model was evaluated using the HDV1 dataset, containing fundus images of different glaucoma stages. Additionally, 40 CNN and 40 ViT models were tested on HDV1 to validate MaxGlaViT's efficiency. Among CNN models, EfficientB6 achieved the highest accuracy (84.91%), while among ViT models, MaxViT-Tiny performed best (86.42%). The scaled MaxViT reached 87.93% accuracy. Adding ECA to the stem block increased accuracy to 89.01%. Replacing MBConv with ConvNeXtV2 further improved it to 89.87%. Finally, integrating ECA in the stem and ConvNeXtV2 in MaxViT blocks resulted in 92.03% accuracy. Testing 80 DL models for glaucoma stage classification, this study presents a comprehensive and comparative analysis. MaxGlaViT outperforms experimental and state-of-the-art models, achieving 92.03% accuracy, 92.33% precision, 92.03% recall, 92.13% f1-score, and 87.12% Cohen's kappa score.

Problem

• Early diagnosis of glaucoma stages
• Lightweight vision transformer model
• Accuracy improvement in glaucoma detection

Innovation

• Lightweight MaxViT architecture optimization
• Enhanced stem with attention mechanisms
• Replaced MBConv with advanced DL blocks