What question did this study set out to answer?

This research aims to create a deep learning pipeline for automated detection and grading of diabetic retinopathy.

July 2, 2026Open Access

Automated diabetic retinopathy grading and screening using deep learning

ARAhmed M. Nashaat Ali RadyCairo University MAMohamed A. AzimBeni-Suef University AAAhmed AbdelMoetySouth Valley University

Key Points

This research aims to create a deep learning pipeline for automated detection and grading of diabetic retinopathy.
Utilized a five-class fundus dataset with 3,500 photographs graded by the ICDR scale.
Employed eight CNN architectures with a 70/20/10 train/validation/test split and transfer learning strategy.
Implemented strong data augmentations and histogram-based contrast enhancements.
VGG16 achieved 76.86% accuracy and the highest Jaccard index for five-class grading.
Binary screening models reached ≥94% accuracy and AUC values between 0.982 and 0.990.
Operating-point calibration supports tailored screening strategies for diverse healthcare systems.

Abstract

Abstract Purpose To develop and benchmark a unified Deep Learning (DL) pipeline for automated detection and five-level grading of Diabetic Retinopathy (DR), and to derive a high-performance binary screening endpoint (DR vs No DR) suitable for scalable use in resource-limited settings. Methods A publicly available five-class DR fundus dataset of 3,500 color photographs graded using the International Clinical Diabetic Retinopathy (ICDR) scale was used. A standardized workflow was applied across eight Convolutional Neural Network (CNN) architectures (AlexNet, Densely Connected Convolutional Network 121 (DenseNet121), Residual Network 50 (ResNet50), eXtreme Inception (Xception), Mobile Network Version 2 (MobileNetV2), Efficient Network Version 2 B2 (EfficientNetV2B2), Inception Version 3 (InceptionV3), Visual Geometry Group 16 (VGG16)), including a 70/20/10 train/validation/test split, optional histogram-based contrast enhancement, strong on-the-fly augmentations, and class-balanced sampling. Seven architectures (DenseNet121, ResNet50, Xception, MobileNetV2, EfficientNetV2B2, InceptionV3, and VGG16) were initialized using ImageNet-pretrained weights and trained using a two-stage transfer-learning strategy with backbone freezing followed by partial fine-tuning, while AlexNet was implemented as a custom architecture and trained from randomly initialized weights. Results For five-class grading, VGG16 achieved the highest accuracy (0.7686) and weighted Jaccard index (0.6572), while EfficientNetV2B2 provided the best balanced accuracy (0.6128) and macro Area Under the Receiver Operating Characteristic Curve (AUROC 0.9158). Misclassifications were concentrated between adjacent severity levels. For binary DR screening, modern backbones achieved ≥0.94 accuracy and balanced accuracy; selected models reached AUROC ≈0.982–0.990 and Area Under the Precision–Recall Curve (AUPRC) ≈ 0.987–0.992. Conclusion The proposed DL pipeline delivers robust multiclass DR grading and highly discriminative binary screening using widely available CNN backbones. Operating-point calibration enables sensitivity-oriented triage or specificity-oriented confirmation, supporting teleophthalmology and task-shifted DR screening programs to expand coverage and reduce preventable vision loss. Clinical trial number Not applicable.

Ask AI

Helpful

Bookmark

View Full Paper