Detection of retinal diseases using an accelerated reused convolutional network

2025-10-05

AI Summary

Deploying deep neural networks for retinal disease detection on resource-constrained devices remains challenging due to the trade-off between accuracy and computational efficiency. To address this, we propose ArConv—a novel lightweight convolutional layer that integrates depthwise separable convolution with structural reuse at the primitive operator level, significantly improving computational efficiency. Based on ArConv, we design an efficient network with only 1.3 million parameters, achieving 93.28% accuracy on the RfMiD dataset—outperforming MobileNetV2 (2.2 million parameters, 92.66% accuracy). This work pioneers the application of ArConv to fundus image-based retinal disease screening, achieving a synergistic breakthrough in model compactness and high accuracy. Our approach establishes a new paradigm for real-time, high-reliability retinal disease screening on mobile platforms.

Technology Category

• Computer Vision: Learning & Optimization for CV
• Machine Learning: Deep Neural Architectures and Foundation Models
• Search and Optimization: Non-convex Optimization

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Convolutional neural networks are continually evolving, with some efforts aimed at improving accuracy, others at increasing speed, and some at enhancing accessibility. Improving accessibility broadens the application of neural networks across a wider range of tasks, including the detection of eye diseases. Early diagnosis of eye diseases and consulting an ophthalmologist can prevent many vision disorders. Given the importance of this issue, various datasets have been collected from the cornea to facilitate the process of making neural network models. However, most of the methods introduced in the past are computationally complex. In this study, we tried to increase the accessibility of deep neural network models. We did this at the most fundamental level, specifically by redesigning and optimizing the convolutional layers. By doing so, we created a new general model that incorporates our novel convolutional layer named ArConv layers. Thanks to the efficient performance of this new layer, the model has suitable complexity for use in mobile phones and can perform the task of diagnosing the presence of disease with high accuracy. The final model we present contains only 1.3 million parameters. In comparison to the MobileNetV2 model, which has 2.2 million parameters, our model demonstrated better accuracy when trained and evaluated on the RfMiD dataset under identical conditions, achieving an accuracy of 0.9328 versus 0.9266 on the RfMiD test set.

Problem

• Detecting retinal diseases using optimized convolutional networks
• Reducing computational complexity for mobile accessibility
• Improving diagnostic accuracy with fewer model parameters

Innovation

• Redesigned convolutional layers for efficiency
• Introduced novel ArConv layers for accessibility
• Optimized model with 1.3 million parameters