Abstract:Color fundus photographs enable the observation of numerous critical biomarkers and early-onset lesions associated with illnesses. Due to its non-invasive and cost-effective nature, this approach can be used for large-scale screening of fundus disorders. In recent years, most applications of data-driven deep learning approaches to fundus illnesses have been based on color fundus photographs. However, current screening methods for fundus illnesses are mainly focused on identifying particular diseases. The majority of published models for multi-disease recognition are ensembles of several binary classification networks, and an ensemble network consumes more resources than a single network. Additionally, many fundus illness recognition models make direct use of established networks for natural picture processing without performing structural optimization, which may reduce disease classification accuracy. To address these issues, we optimized the network structure to build a single multi-label fundus disease recognition model. Specifically, we began by selecting EfficientNet-B4 as a backbone network, then modified the output layer to create a multi-label recognition model. Based on this model, the spatial attention structure was extended to improve the network’s feature extraction capability. The focal loss was used to increase classification accuracy by mitigating the training effects of data imbalance. Our proposed model significantly improves recognition performance on a publicly available dataset compared to other baseline networks. When spatial attention and focal loss function are introduced into the backbone model, the baseline values of all relevant evaluation indicators are improved without significant increase in computational cost. Additionally, this article provides two error correction strategies for multi-label classification issues: mutual exclusion and super label space. When a more robust error correction strategy is employed with the model, F1 increases by 2.86% in comparison to the original performance of EfficientNet-B4.

A Single Network with an Attention Mechanism for Glaucoma Detection

Associations between self-assessed masticatory disability and health of community-residing elderly people.

Attention Mechanism-Based Glaucoma Classification Model Using Retinal Fundus Images

Optic disc and cup segmentation for glaucoma detection using Attention U-Net incorporating residual mechanism

Automated glaucoma screening method based on image segmentation and feature extraction

AlterNet-K: a small and compact model for the detection of glaucoma

Attention Based Glaucoma Detection: A Large-scale Database and CNN Model

Multi-GlaucNet: A multi-task model for optic disc segmentation, blood vessel segmentation and glaucoma detection

An Artificial-Intelligence- and Telemedicine-Based Screening Tool to Identify Glaucoma Suspects from Color Fundus Imaging

Detection of glaucoma using three-stage training with EfficientNet

A Large-Scale Database and a CNN Model for Attention-Based Glaucoma Detection

Self-Adaptive Transfer Learning for Multicenter Glaucoma Classification in Fundus Retina Images

Regression and Learning with Pixel-wise Attention for Retinal Fundus Glaucoma Segmentation and Detection

GLA-Net: A global-local attention network for automatic cataract classification

EfficientNet embedded with spatial attention for recognition of multi-label fundus disease from color fundus photographs

GARDNet: Robust Multi-View Network for Glaucoma Classification in Color Fundus Images

En‐ConvNet: A novel approach for glaucoma detection from color fundus images using ensemble of deep convolutional neural networks

Glaucoma disease detection using stacked attention U-Net and deep convolutional neural network

Computerized Assessment of Glaucoma Severity Based on Color Fundus Images

GAMNet: a gated attention mechanism network for grading myopic traction maculopathy in OCT images

GLIM-Net: Chronic Glaucoma Forecast Transformer for Irregularly Sampled Sequential Fundus Images