Rama Devi C, Rithanya K.R, Shakti P, Subhashini G | International Journal of Image Processing and Pattern Recognition | Vol 12, Issue 1 | pp. 8-16 | ISSN: 2456-6985
Abstract
Diabetic retinopathy (DR) is a leading cause of blindness worldwide, impacting the largest number of individuals who have diabetes. Early identification of diabetic retinopathy may help to prevent severe effects; however, the very limited availability of labeled datasets, and the possibility of overfitting of the model, presents obstacles to obtaining an accurate diagnosis. In this research, a hybrid deep learning strategy is introduced, where Generative Adversarial Networks (GANs) are used for data augmentation and ResNet-50-based Convolutional Neural Networks (CNNs) for DR classification. The GANs produce artificial retinal images, promoting dataset diversity and overfitting prevention, while the ResNet-50 model is optimized for effective feature extraction from fundus images. The objective is to enhance the accuracy and sensitivity of detection of DR. The system proposed performs better than conventional CNNs and other models such as VGG16 and InceptionV3, with an accuracy of 86.7% in the Kaggle DR dataset. The benefits of the system are improved performance, less overfitting, and the capability for early and effective DR detection that could result in improved clinical results.
Keywords
early detection, Convolutional Neural Networks, diabetic retinopathy, fundus images, data augmentation, generative adversarial networks
🔒 This is a subscription article
Full text is available to subscribers and institutional members. Please choose an option below to access it.
SubscribePurchase this articleInstitutional / Login accessReferences
- Atwany MZ, Sahyoun AH, Yaqub M. Deep learning techniques for diabetic retinopathy classification: A survey. IEEE Access. 2022;10:28642–28655.
- Zhu S, Xiong C, Zhong Q, Yao Y. Diabetic retinopathy classification with deep learning via fundus images: A short survey. IEEE Access. 2024;12:20540–20558.
- Gao J, Li S, Chen Y, Xiang R. MSAmix-net: Diabetic retinopathy classification. IEEE Access. 2024 Nov 26.
- Rajarajeshwari G, Selvi GC. Application of artificial intelligence for classification, segmentation, early detection, early diagnosis, and grading of diabetic retinopathy from fundus retinal images: A comprehensive review. IEEE Access. 2024 Nov 11.
- Jagadesh BN, Karthik MG, Siri D, Shareef SK, Mantena SV, Vatambeti R. Segmentation using the IC2T model and classification of diabetic retinopathy using the rock hyrax swarm-based coordination attention mechanism. IEEE Access. 2023;11:124441–124458.
- Kadhim AJ, Seyedarabi H, Afrouzian R, Hasan FS. Diabetic retinopathy classification using hybrid color-based CLAHE and blood vessel in deep convolution neural network. IEEE Access. 2024 Dec 17.
- Zedadra A, Zedadra O, Salah-Salah MY, Guerrieri A. Graph-aware multimodal deep learning for classification of diabetic retinopathy images. IEEE Access. 2025 Apr 25.
- Nazih W, Aseeri AO, Atallah OY, El-Sappagh S. Vision transformer model for predicting the severity of diabetic retinopathy in fundus photography-based retina images. IEEE Access. 2023;11:117546–117561.
- Majumder S, Kehtarnavaz N. Multitasking deep learning model for detection of five stages of diabetic retinopathy. IEEE Access. 2021;9:123220–123230.
- Tariq M, Palade V, Ma Y. Effective diabetic retinopathy classification with siamese neural network: A strategy for small dataset challenges. IEEE Access. 2024 Dec 2.
- Alavee KA, Hasan M, Zillanee AH, Mostakim M, Uddin J, Alvarado ES, et al. Enhancing early detection of diabetic retinopathy through the integration of deep learning models and explainable artificial intelligence. IEEE Access. 2024;12:73950–73969.
- Infantia Henry N, Anbuananth C, Kalarani S. Hybrid meta-heuristic algorithm for optimal virtual machine placement and migration in cloud computing. Concurrency Comput Pract Exp. 2022;34(28):e7353.
- Infantia Henry N, Anbuananth C, Kalarani S. An efficient and robust method for data privacy and security on a public cloud using a novel hybrid technique. Eng Proc. 2023;59(1):115.
- Subhashini G, Chandrasekar A. Hybrid deep learning technique for optimal segmentation and classification of multi-class skin cancer. Imaging Sci J. 2024;72(8):1043–1064.
- Subhashini G, Devi A, Infantia CN, Karthi M, Raghul G, Shankar SA. U-NET and RCNN ensembled satellite object detection and segmentation. In: Proc Third Int Conf Artif Intell Smart Energy (ICAIS); 2023 Feb 2; India. IEEE. p. 836–844.
- Subhashini G, Devi CR, Rasool SB. Eyesight impairment detection for diabetic patients using very deep convolution neural network for 3-channel retinal images. In: Proc 6th Int Conf Trends Electron Informatics (ICOEI); 2022 Apr 28. IEEE. p. 1129–1134.
- Patra R, Khuntia B, Panda DC. Fractional rider gradient descent applied U-net based segmentation with optimal deep maxout network for lung cancer classification using histopathological images. Res Biomed Eng. 2022;38(2):599–615.
How to cite this article
@article{CRD2026,
author = {Rama Devi C and Rithanya K.R and Shakti P and Subhashini G},
title = {Advanced Classification of Diabetic Retinopathy Using GAN and Deep Residual CNN Model},
journal = {International Journal of Image Processing and Pattern Recognition},
year = {2026},
volume = {12},
number = {1},
pages = {8--16},
issn = {2456-6985},
url = {https://journalspub.com/publication/ijippr/article=26325}
}