Convolutional Neural Networks for Diabetic Retinopathy Detection in Retinal Images
Keywords:
Diabetic retinopathy, Ocular ailment, Uncontrolled diabetes, Elevated blood, sugar, Convolutional Neural Networks (CNN), Deep learning models, Medical imagingAbstract
Diabetic retinopathy is a degenerative eye disease associated with uncontrolled diabetes, high blood pressure, blood sugar, cholesterol, and body weight. If diabetes-related diabetic retinopathy is not detected and treated in its early stages, it might result in visual impairment. Traditional methods of diagnosis often require manual examination by trained ophthalmologists, which can be time-consuming and subjective. Deep learning models called Convolutional Neural Networks (CNNs) are renowned for their capacity to recognize small patterns and features in images, which makes them very useful for challenging medical imaging applications. In this study, we proposed two CNN architectures, namely CNN-Plain and CNN-BN-D to automate and enhance the diagnostic process for diabetic retinopathy (DR) detection. The model's effectiveness is assessed using a range of criteria, such as accuracy, precision, recall, and F1Score, to confirm that it is as effective as current diagnostic techniques. The results demonstrate the CNN-BN-D model with 0.94 accuracy on train and test data exhibits superior performance and generalization compared to the simpler CNN-Plain architecture with 0.89 accuracy in the task of diabetic retinopathy detection.
Downloads
References
S. Rykov, D. Chugaiev, and S. Ziablitsev, “Blood selectin levels as a predictive fac- tor for diabetic retinopathy and diabetic macular edema in type 2 diabetes.” Journal of Ophthalmology (Ukraine)/Oftalmologicˇeskij Žurnal, no. 3, 2023.
N. u. Huda, A. A. Salam, N. S. Alghamdi, J. Zeb, and M. U. Akram, “Prolifera- tive diabetic retinopathy diagnosis using varying-scales filter banks and double- layered thresholding,” Diagnostics, vol. 13, no. 13, p. 2231, 2023.
A. Shanthini, G. Manogaran, G. Vadivu, K. Kottilingam,P. Nithyakani, and C. Fancy, “Threshold segmentation based multi-layer analysis for detecting diabetic retinopathy using convolution neural net- work,” Journal of Ambient Intelligence and Humanized Computing, pp. 1–15, 2021.
S. Burewar, A. B. Gonde, and S. K. Vip- parthi, “Diabetic retinopathy detection by retinal segmentation with region merging using cnn,” in 2018 IEEE 13th International Conference on Industrial and Information Sys- tems (ICIIS). IEEE, 2018, pp. 136–142.
P. H. N. Castro, G. C. Fortuna, R. A. B. de Queiroz, and G. J. P. Moreira, “Regu- larization through simultaneous learning: A case study for hop classification,” arXiv preprint arXiv:2305.13447, 2023.
R. Vignesh, N. Muthukumaran, and M. P. Austin, “Detection of diabetic retinopathy image analysis using convolution graph neural network,” in 2023 International Con- ference on Inventive Computation Technolo- gies (ICICT). IEEE, 2023, pp. 921–929.
K. Swarnalatha, U. A. Nayak, N. A. Benny,H. Bharath, D. Shetty, and S. D. Kumar, “Detection of diabetic retinopathy using convolution neural network,” in Emerging Research in Computing, Information, Com- munication and Applications: Proceedings of ERCICA 2022. Springer, 2022, pp. 427–439.
D. Mane, S. Sangve, P. Kumbharkar, S. Rat- naparkhi, G. Upadhye, and S. Borde, “A diabetic retinopathy detection using cus- tomized convolutional neural network.”
X. Shen and J. Lin, “Consistency of neu- ral networks with regularization,” arXiv preprint arXiv:2207.01538, 2022.
M. Xu, T. Zhang, Z. Li, and D. Zhang, “Im- proving the certified robustness of neural networks via consistency regularization,” arXiv preprint arXiv:2012.13103, 2020.
P. Sinha, I. Psaromiligkos, and Z. Zilic, “A structurally regularized cnn architec- ture via adaptive subband decomposition,” arXiv preprint arXiv:2306.16604, 2023.
M. M. Farag, “Design and analysis of con- volutional neural layers: A signal process- ing perspective,” IEEE Access, vol. 11, pp. 27 641–27 661, 2023.
G. Michelogiannakis, X. S. Li, D. H. Bailey, and J. Shalf, “Extending summation preci- sion for network reduction operations,” In- ternational Journal of Parallel Programming, vol. 43, pp. 1218–1243, 2015.
C. Chen, F. Min, Y. Zhang, and H. Bao, “Relu-type hopfield neural network with analog hardware implementation,” Chaos, Solitons & Fractals, vol. 167, p. 113068, 2023.
M. Monnet, H. Gebran, A. Matic-Flierl, F. Kiwit, B. Schachtner, A. Bentellis, and J. M. Lorenz, “Pooling techniques in hybrid quantum-classical convolu- tional neural networks,” arXiv preprint arXiv:2305.05603, 2023.
S. M. Ko, S. Cho, D.-W. Jeong, S. Han, M. Lee, and H. Lee, “Grouping matrix based graph pooling with adaptive num- ber of clusters,” in Proceedings of the AAAI Conference on Artificial Intelligence, vol. 37, no. 7, 2023, pp. 8334–8342.
D. Sun, J. Wulff, E. B. Sudderth, H. Pfis- ter, and M. J. Black, “A fully-connected layered model of foreground and back- ground flow,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2013, pp. 2451–2458.
C. Zhao, G. Dong, S. Zhang, Z. Tan, and A. Basu, “Frequency regularization: Restricting information redundancy of convolutional neural networks,” arXiv preprint arXiv:2304.07973, 2023.
Z. Liu, Z. Xu, J. Jin, Z. Shen, and T. Darrell, “Dropout reduces underfitting,” arXiv preprint arXiv:2303.01500, 2023.
Kim, Bum Jun, Hyeyeon Choi, Hyeonah Jang, Donggeon Lee, and Sang Woo Kim. "How to Use Dropout Correctly on Residual Networks with Batch Normalization." arXiv preprint arXiv:2302.06112 (2023).
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
All papers should be submitted electronically. All submitted manuscripts must be original work that is not under submission at another journal or under consideration for publication in another form, such as a monograph or chapter of a book. Authors of submitted papers are obligated not to submit their paper for publication elsewhere until an editorial decision is rendered on their submission. Further, authors of accepted papers are prohibited from publishing the results in other publications that appear before the paper is published in the Journal unless they receive approval for doing so from the Editor-In-Chief.
IJISAE open access articles are licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. This license lets the audience to give appropriate credit, provide a link to the license, and indicate if changes were made and if they remix, transform, or build upon the material, they must distribute contributions under the same license as the original.