DeepRetina: Detection of Diabetic Retinopathy

In this work, we have proposed an automated DeepRetina system for detecting and classifying diabetic retinopathy using deep learning. This method involves Convolutional Neural Networks, a class of deep learning algorithm for analyzing retinal fundus images and predicting the stages of the disease, utilizing transfer learning. It also involves preprocessing, data augmentation, and model tuning to achieve better classification. While traditional machine learning techniques depend on hand-crafted features to extract image characteristics, deep learning techniques are more efficient in learning complex features in the image data, hence yielding better accuracy.   In our experiments, the proposed method produced reliable results compared to other research work in this area. The use of CNN based model is encouraged and widespread for feature extraction from image data for disease detection tasks yielding significant accuracy in the analysis of the disease. Transfer learning technique provides better generalization of model and significantly reduces the computational cost of the training process, making it suitable for practical deployment. Explainable AI further assists to understand the decision made by the model through visualization of important features of image that leads to final prediction.   Therefore, we have developed a system called DeepRetina which would effectively aid the physicians in quick and timely screening of diabetic retinopathy and help to prevent vision impairment. By performing these tasks effectively and efficiently the workload can be significantly reduced, and eventually better results could be attained in this field by large scale implementation.    

1.        Sindhu Madhuri G., Chokkanathan K., Mahesh T. R., et al., “MLPDR: High Performance ML Algorithms for the Prediction of Diabetic Retinopathy,” 2023 International Conference on Network, Multimedia and Information Technology (NMITCON), IEEE, 2023.   2.        Wong, W. K., Juwono, F. H., and Apriono, C., “Diabetic Retinopathy Detection and Grading: A Transfer Learning Approach Using Simultaneous Parameter Optimization and Feature-Weighted ECOC Ensemble,” IEEE Access, vol. 11, pp. 83004–83006, 2023.   3.        Chandra, R., Tiwari, S., Kumar, S. S., and Agarwal, S., “Diabetic Retinopathy Prediction based on CNN and AlexNet Model,” 2024 International Conference on Cloud Computing, Data Science & Engineering (Confluence), IEEE, 2024.   4.        Nor Azamen, N. T. N., Mohd Ali, A., and Abd Aziz, N. A., “Prediction of Diabetic Retinopathy Based on Risk Factors using Machine Learning Algorithms,” 2023 4th International Conference on Artificial Intelligence and Data Sciences (AiDAS), IEEE, 2023.   5.        Bhatia, K., Arora, S., and Tomar, R., “Diagnosis of Diabetic Retinopathy Using Machine Learning Classification Algorithm,” 2016 2nd International Conference on Next Generation Computing Technologies (NGCT), IEEE, 2016.   6.        Gulshan, V., Peng, L., Coram, M., et al., “Development and Validation of a Deep Learning Algorithm for Detection of Diabetic Retinopathy,” Journal of the American Medical Association (JAMA), vol. 316, no. 22, pp. 2402–2410, 2016.   7.        Pratt, H., Coenen, F., Broadbent, D. M., Harding, S. P., and Zheng, Y., “Convolutional Neural Networks for Diabetic Retinopathy,” Procedia Computer Science, vol. 90, pp. 200–205, 2016.   8.        Ting, D. S. W., Cheung, C. Y., Lim, G., et al., “Development and Validation of a Deep Learning System for Diabetic Retinopathy and Related Eye Diseases,” Nature Medicine, vol. 25, pp. 206–212, 2019.   9.        Abramoff, M. D., Lavin, P. T., Birch, M., Shah, N., and Folk, J. C., “Pivotal Trial of an Autonomous AI-Based Diagnostic System for Detection of Diabetic Retinopathy,” NPJ Digital Medicine, vol. 1, no. 39, 2018.