Cataract Classification in Eye Images Using MobileNetV2
Abstract views: 16 , PDF downloads: 11Abstract
Cataract remains one of the primary causes of visual impairment globally, with early detection being essential to prevent permanent blindness and improve patient quality of life. However, conventional diagnosis depends on ophthalmologists and clinical-grade imaging devices, which are often limited in remote or under-resourced areas. This condition highlights the need for an efficient, accessible, and automated screening solution. To address this challenge, this study utilizes the MobileNetV2 deep learning architecture to classify cataract conditions based on eye images. MobileNetV2 is selected because of its lightweight model structure and strong feature representation capabilities, making it suitable for deployment in portable or embedded medical systems. The dataset used consists of two cataract stages, namely immature and mature cataracts, with images undergoing preprocessing prior to model training. The proposed system demonstrates excellent performance, achieving an accuracy, precision, recall, and F1-score of 100% in distinguishing cataract stages. These results confirm that MobileNetV2 can effectively support cataract screening with high reliability while maintaining efficiency. Future work will involve extending the dataset to include additional cataract severity levels and non-cataract eye images, as well as integrating explainable artificial intelligence methods to provide visual diagnostic interpretations and enhance clinical trust in real-world applications.
Downloads
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
References
[2] M. Abou Shousha et al., “Automated nuclear cataract grading using anterior segment OCT and deep learning,” Ophthalmology Science, vol. 1, no. 2, pp. 100051, 2021.
[3] Y. Chen, K. Li, and L. Liu, “MobileNet-based cataract detection for portable medical screening systems,” Biomedical Signal Processing and Control, vol. 66, pp. 102493, 2021.
[4] S. Xu, H. Hu, and Q. Wang, “Tele-ophthalmology supported by AI for cataract screening in rural populations,” Frontiers in Medicine, vol. 8, pp. 742573, 2021.
[5] A. Bhat and S. Shankar, “Lightweight CNN with LBP for cataract staging in low-resource settings,” Computer Methods and Programs in Biomedicine, vol. 212, pp. 106460, 2021.
[6] R. Mittal, P. Singh, and K. Verma, “Optimized deep convolutional networks for cataract severity classification,” Computers in Biology and Medicine, vol. 134, pp. 104485, 2021.
[7] S. Alam, T. Islam, and M. Khan, “Recent progress in computer-aided cataract diagnosis: A systematic review,” Pattern Recognition Letters, vol. 152, pp. 175–185, 2021.
[8] Y. Zhang et al., “Real-time cataract detection on embedded devices using edge AI,” IEEE Internet of Things Journal, vol. 9, no. 7, pp. 5510–5520, 2022.
[9] W. Li, J. Zhao, and S. Chen, “Transfer learning for anterior eye disease classification in ophthalmology,” Applied Sciences, vol. 12, no. 3, pp. 1345, 2022.
[10] K. Rahman and S. Islam, “Explainable deep learning for cataract diagnosis: Enhancing clinical interpretability,” Diagnostics, vol. 12, no. 8, pp. 1910, 2022.
[11] J. Sun, L. Zhou, and F. Li, “Deep feature-based cataract severity grading using slit-lamp images,” BMC Ophthalmology, vol. 22, no. 345, 2022.
[12] A. D. Santoso, N. A. Nugroho, and R. F. Rahman, “Mobile-based cataract detection using CNN in telemedicine services,” Indonesian Journal of Electrical Engineering and Informatics, vol. 10, no. 4, pp. 892–902, 2022.
[13] World Health Organization (WHO), “World Report on Vision: Global Update,” Geneva: WHO Press, 2023.
[14] International Agency for the Prevention of Blindness (IAPB), “Vision Atlas: Global burden of cataract,” 2024.
[15] L. Song et al., “Multi-stage cataract classification using attention-guided convolutional networks,” Scientific Reports, vol. 13, no. 1884, 2023.
[16] Z. Huang, F. Wei, and D. Tan, “Cross-device generalization of deep-learning cataract classifiers,” IEEE Transactions on Medical Imaging, vol. 43, no. 1, pp. 112–123, 2024.
[17] P. Zhang and H. Luo, “Improving cataract detection with hybrid CNN-transformer architectures,” Artificial Intelligence in Medicine, vol. 145, pp. 102676, 2024.
[18] T. Nguyen, J. Park, and M. Lee, “Efficient lightweight CNNs for medical image diagnosis on mobile devices,” IEEE Access, vol. 12, pp. 88245–88259, 2024.
[19] F. Wang et al., “Deep learning for analyzing crystalline lens opacity progression,” Eye and Vision, vol. 11, no. 2, pp. 25–37, 2024.
[20] R. Arrioja, L. Flores, and S. Delgado, “Telehealth-based cataract screening in rural communities: Implementation outcomes,” Journal of Telemedicine and Telecare, vol. 31, no. 1, pp. 77–89, 2024.
[21] A. Howard et al., “MobileNetV2: Inverted residuals and linear bottlenecks (revised reproducibility release),” IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2022.
[22] R. Lin, Y. He, and L. Qiu, “Explainable AI-based assistance for ophthalmic diagnosis,” IEEE Reviews in Biomedical Engineering, vol. 18, pp. 101–118, 2025.
[23] H. Tamura and M. Okada, “AI-assisted cataract triage in national health screening programs,” Lancet Digital Health, vol. 7, no. 1, pp. e14–e24, 2025.
