Identification of Eye Diseases from Fundus Images Using Convolutional Neural Network with ResNet50 Architecture
Article Sidebar
Main Article Content
Abstract
This study aims to identify eye diseases from fundus images using ResNet50-based Convolutional Neural Network (CNN) architecture with a transfer learning approach. The dataset used comes from Kaggle with a total of 4217 images, covering four classes of eye diseases: Diabetic Retinopathy, Glaucoma, Cataract, and Normal. The process includes preprocessing with augmentation and normalization, transfer learning using a pre-trained ResNet50 model on imagesNet, and evaluation with a confusion matrix. The results show a testing accuracy of 87.91%, with the best performance in the Diabetic Retinopathy class and challenges in the Glaucoma class. Suggestions include balancing the dataset and further fine-tuning.
Article Details
This work is licensed under a Creative Commons Attribution 4.0 International License.
References
[1] M. S. Qulub and S. Agustin, “Indentifikasi Penyakit Mata Dengan Klasifikasi Citra Foto Fundus Mengunakan Convolutional Neural Network (CNN),” vol. 8, no. 5, pp. 11 034–11 039, 2024.
[2] Cahya and Suwanda, “Implementasi augmentasi data dalam klasifikasi penyakit mata menggunakan cnn,” Jurnal Teknologi Informasi dan Ilmu Komputer, vol. 6, no. 2, pp. 115–122, 2021.
[3] D. Juniati and A. E. Suwanda, “Klasifikasi Penyakit Mata Berdasarkan Citra Fundus Retina Menggunakan Dimensi Fraktal Box Counting Dan Fuzzy K-Means,” Proximal: Jurnal Penelitian Matematika dan Pendidikan Matematika, vol. 5, no. 1, pp. 10–18, 2022.
[4] A. Verdy and S. Hartati, “High-accuracy eye disease classification using resnet-50 architecture,” International Journal of Medical Imaging, vol. 12, no. 2, pp. 55–63, 2024.
[5] Y. Huang, C. Xu, H. Xu, dan Z. Fang, “Identifying key components in ResNet-50 for diabetic retinopathy grading and model deployment,” arXiv preprint arXiv:2110.14160, 2021.
[6] J. Puchaicela-Lozano, A. Guerrero-Curieses, F. Rodriguez, dan J. I. Arribas, “Glaucoma detection on fundus images using Faster R-CNN and transfer learning,” Proceedings of the 19th International Conference on Informatics in Control, Automation and Robotics (ICINCO), 2023, pp. 338–345.
[7] H. Abdel-Aty, M. Elhoseny, M. Zaki, dan M. Abd Elaziz, “Deep learning approach for fundus image classification using Inception-v3 and ResNet-50,” Frontiers in Physiology, vol. 14, art. no. 1126780, 2023.
[8] J. Deng, X. Wu, dan L. Zhang, “A revised ResNet-50 for diabetic retinopathy detection with improved generalization,” BMC Bioinformatics, vol. 24, no. 1, art. no. 153, 2023.
[9] R. Mishra, A. Srivastava, dan D. Sharma, “Detection of diabetic retinopathy using VGG19 and ResNet-50 models,” ResearchGate Preprint, 2024.
[10] H. Aly, M. F. Elkhateeb, dan A. M. Soliman, “Glaucoma detection using ensemble deep learning models and Grad-CAM visualization,” Arab Journal of Basic and Applied Sciences, vol. 31, no. 1, pp. 1–12, 2024.