Diabetic Retinopathy Prediction Using Deep Learning: Insights From CNN
Article Sidebar
Deep Learning,
Diabetic Retinopathy,
Medical Imaging,
SHAP,
Transfer Learning
Main Article Content
Ranga Swamy Sirisati
V. Navyasri
T. Swathi
M. Akhila
Astried Astried
Abstract
Background of study: Diabetic Retinopathy (DR) is a severe microvascular complication of diabetes mellitus that can lead to vision loss if not detected early. With over 93 million individuals affected globally, the need for accurate and efficient diagnostic systems has become urgent. Traditional screening methods depend on manual interpretation of fundus images by ophthalmologists, which is time-consuming and prone to subjectivity.
Aims: This research seeks to create and assess a deep learning diagnostic model designed to reliably identify the severity levels of diabetic retinopathy using retinal fundus images. The research also explores model interpretability using Shapley Additive Explanations (SHAP) to increase transparency in AI-assisted medical diagnosis.
Methods: Convolutional Neural Networks (CNNs) were implemented using transfer learning with pretrained architectures such as ResNet50 and InceptionV3. The EyePACS dataset, containing images categorized into five DR severity levels, was used for model training. Preprocessing techniques, including contrast enhancement, histogram equalization, and data augmentation, improved image quality and model generalization. The models were optimized with the Adam and assessed through accuracy, precision, recall, F1-score, and AUC. Additionally, SHAP analysis was employed to interpret and illustrate the model’s predictions.
Results: The proposed CNN-based model achieved 98.5% accuracy, with a sensitivity and specificity of 0.99, demonstrating strong performance across multiple DR stages. Comparison with existing studies revealed a notable improvement in diagnostic accuracy. SHAP visualizations confirmed that critical retinal features such as microaneurysms, hemorrhages, and cotton-wool spots were key predictors influencing model decisions.
Conclusion: The findings validate the efficacy of deep learning, particularly CNNs, in enhancing early detection and classification of diabetic retinopathy. The integration of SHAP interpretability bridges the gap between AI predictions and clinical trust, making this approach a promising tool for large-scale automated DR screening and supporting ophthalmologists in timely diagnosis and treatment.
Article Details
Copyright (c) 2025 Ranga Swamy Sirisati, V. Navyasri, T. Swathi, M. Akhila, Astried Astried
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
References
Abdullah S. Canipek. (2023). Eyepacs, Aptos, Messidor Diabetic Retinopathy. https://www.kaggle.com/ datasets/ascanipek/eyepacs-aptos-messidor-diabetic-retinopathy
Agrawal, B., & Panicker, S. (2024). Advancements in Diabetic Retinopathy Detection : Analyzing the Efficacy of Supervised Learning Algorithms. Journal of Cardiovascular Disease Research, 15(11), 481–489. https://doi.org/10.48047/
Aldughayfiq, B., Ashfaq, F., & Jhanjhi, N. Z. (2023). Explainable AI for Retinoblastoma Diagnosis : Interpreting Deep Learning Models with LIME and SHAP. Diagnostics, 13(11), 1–19. https://doi.org/10.3390/diagnostics13111932
Ali, Q., Heldal, I., Helgesen, C. G., Krumina, G., Costescu, C., Kovari, A., Katona, J., & Thill, S. (2021). applied sciences Current Challenges Supporting School-Aged Children with Vision Problems : A Rapid Review. Applied Sciences, 11(20), 1–23. https://doi.org/10.3390/app11209673
Alkhouly, A., Mohammed, A., & Hefny, H. (2021). Improving The Performance Of Deep Neural Networks Using Two Proposed Activation Functions. IEEE Access, 4, 82249–82271. https://doi.org/10.1109/ACCESS.2021.3085855
Ansari, P., Tabasumma, N., Snigdha, N. N., & Siam, N. H. (2022). Diabetic Retinopathy : An Overview on Mechanisms , Pathophysiology and Pharmacotherapy. Diabetology, 3(1), 159–175. https://doi.org/10.3390/diabetology3010011
Batista, T. M., Haider, N., & Kahn, C. R. (2021). Defining the underlying defect in insulin action in type 2 diabetes. Diabetologia, 64(5), 994–1006. https://doi.org/10.1007/s00125-021-05415-5
Chaudhary, S., Zaveri, J., & Becker, N. (2021). Disease-a-Month Proliferative diabetic retinopathy ( PDR ). Disease-a-Month, 67(5), 101140. https://doi.org/10.1016/j.disamonth.2021.101140
Gettinger, K., Lee, D., Tomita, Y., & Negishi, K. (2025). Diabetic Retinopathy , a Comprehensive Overview on Pathophysiology and Relevant Experimental Models. International Journal of Molecular Sciences, 26(20), 1–25. https://doi.org/10.3390/ijms26209882
Gomułka, K., & Ruta, M. (2023). The Role of Inflammation and Therapeutic Concepts in Diabetic Retinopathy — A Short Review. International Journal of Molecular Sciences, 24(3), pag 1024. https://doi.org/10.3390/ijms24021024
Gregg, E., Buckley, J., Ali, M. K., Davies, J., Flood, D., Mehta, R., Griffiths, B., Lim, L., Manne-goehler, J., Pearson-, J., Tandon, N., & Shaw, J. E. (2023). Improving health outcomes of people with diabetes: target setting for the WHO Global Diabetes Compact. HHS Public Access, 401(10384), 1302–1312. https://doi.org/10.1016/S0140-6736(23)00001-6.Improving
Hannan, A., Mahmood, Z., Qureshi, R., & Ali, H. (2025). Computer Methods in Biomechanics and Biomedical Engineering : Imaging & Visualization Enhancing diabetic retinopathy classification accuracy through dual-attention mechanism in deep learning. Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 13(1). https://doi.org/10.1080/21681163.2025.2539079
He, L., Luan, L., & Hu, D. (2025). Deep learning-based image classification for AI-assisted integration of pathology and radiology in medical imaging. Frontiers in Medicine, 2(12). https://doi.org/10.3389/fmed.2025.1574514
Huang, X., Wang, H., She, C., & Feng, J. (2022). Arti fi cial intelligence promotes the diagnosis and screening of diabetic retinopathy. Frontiers in Endocrinology, 13(September), 1–12. https://doi.org/10.3389/fendo.2022.946915
Li, Z., Liu, F., Yang, W., Peng, S., Zhou, J., & Member, S. (2021). A Survey of Convolutional Neural Networks : Analysis , Applications , and Prospects. IEEE Transactions on Neural Networks and Learning Systems, 33(12), 1–21. https://doi.org/10.1109/TNNLS.2021.3084827
Liu, Y., Pu, H., & Sun, D. (2021). Efficient extraction of deep image features using convolutional neural network ( CNN ) for applications in detecting and analysing complex food matrices. Trends in Food Science & Technology, 113(October 2020), 193–204. https://doi.org/10.1016/j.tifs.2021.04.042
Mienye, I. D., Swart, T. G., Obaido, G., Jordan, M., & Ilono, P. (2025). Deep Convolutional Neural Networks in Medical Image Analysis : A Review. Information, 16(3), 1–28. https://doi.org/10.3390/info16030195
Muhathir, M., Ryandra, M. F. D., Syah, R. B. Y., & Khairina, N. (2023). Convolutional Neural Network (CNN) of Resnet-50 with Inceptionv3 Architecture in Classification on X-Ray Image. Lecture Notes in Networks and SystemsLecture Notes in Networks and Systems. https://doi.org/10.1007/978-3-031-35314-7_20
Olawade, D. B., Weerasinghe, K., Don, M., Eranga, D., Odetayo, A., Aderinto, N., Teke, J., & Boussios, S. (2025). Enhancing Ophthalmic Diagnosis and Treatment with Artificial Intelligence. Medicina, 61(3), 1–23. https://doi.org/10.3390/medicina61030433
Pinto-coelho, L. (2023). How Artificial Intelligence Is Shaping Medical Imaging Technology : A Survey of Innovations and Applications. Bioengineering, 10(12), 1435. https://doi.org/10.3390/bioengineering10121435
Rêgo, S., Monteiro-Soares, M., Soares, F., & Medeiros, M. D. (2022). Interobserver variability in the determination of diabetic retinopathy and quality of fundus image. Photo Acoustic and Optical Coherence Tomography Imaging. https://doi.org/10.1088/978-0-7503-2056-6ch5?urlappend=%3Futm_source%3Dresearchgate.net%26utm_medium%3Darticle
Salehi, A. W., Khan, S., Gupta, G., Alabduallah, B. I., & Almjally, A. (2023). A Study of CNN and Transfer Learning in Medical Imaging : Advantages , Challenges , Future Scope. Sustainability, 15(7), Page 5930. https://doi.org/10.3390/su15075930
Skevas, C., Olaguer, N. P. De, Lleó, A., Thiwa, D., Schroeter, U., & Lopes, I. V. (2024). Implementing and evaluating a fully functional AI-enabled model for chronic eye disease screening in a real clinical environment. BMC Ophthalmology, 1–12. https://doi.org/10.1186/s12886-024-03306-y
Syaifudin, T. S., Nurjanah, M. H., Kumalasari, C., Widodo, W. T., Studi, P., & Kesehatan, A. (2022). RELATIONSHIP BETWEEN HbA1C AND eGFR IN DIABETES MELLITUS ( DM ) PATIENTS FOLLOWING PROLANIS AT ULTRA. Jurnal Biosains Pascasarjana, 24(1), 13–20. https://doi.org/10.20473/jbp.v24i1SP.2022.13-20
Wei, L., Sun, X., Fan, C., Li, R., Zhou, S., Yu, H., & Miller, W. (2022). The pathophysiological mechanisms underlying diabetic retinopathy. Frontiers in Cell and Developmental Biology, 10(August), 1–12. https://doi.org/10.3389/fcell.2022.963615
Zangana, H. M., Mohammed, A., & Mustafa, F. M. (2024). Advancements and Applications of Convolutional Neural Networks in Image Analysis : A Comprehensive Review. JURNAL ILMIAH COMPUTER SCIENCE (JICS), 3(1), 16–29. https://doi.org/10.58602/jics.v3i1.30