Abstract

Cataract is a serious eye disease which may cause blindness. Early detection is of high significance to the treatment of cataract, which reduces the risk of patients to turning into blindness. Some studies were conducted for fundus image classification based on traditional machine learning methods. However, their performance still can be improved. Therefore, a novel deep learning based method is proposed to classify cataract using fundus images. To avoid relying on mass labelled data, the proposed method resorts to deep transfer learning to reduce the number of parameters that need to be trained. Consequently, in the proposed method, the first five convolutional layers of AlexNet are utilized for general feature learning; then three subsequent layers are designed and fine-tuned for the classification of fundus images. The best performance for cataract detection is 95.37% in terms of classification accuracy.

References

• Miguel Caixinha, Elena Velte, M´ario Santos, and Jaime B Santos. New approach for objective cataract classification based on ultrasound techniques using multiclass svm classifiers. In Ultrasonics Symposium (IUS), 2014 IEEE International, pages 2402–2405. IEEE, 2014. Proposed: 95.37%Model: 90.90%Model: 93.20% Model: 95.22% 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Accuracy Fig. 4: Performance comparison with two related models.
• Jia Deng, Wei Dong, R. Socher, Li Jia Li, Kai Li, and Fei Fei Li. Imagenet: A large-scale hierarchical image database. Proc of IEEE Computer Vision & Pattern Recognition, pages 248– 255, 2009.
• Weiming Fan, Ruifang Shen, Qinyan Zhang, Ji-Jiang Yang, and Jianqiang Li. Principal component analysis based cataract grading and classification. In E-health Networking, Application & Services (HealthCom), 2015 17th International Conference on, pages 459–462. IEEE, 2015.
• Allen Foster et al. Vision 2020: the cataract challenge. Community Eye Health, 13(34):17–19, 2000.
• Xinting Gao, Huiqi Li, Joo Hwee Lim, and Tien Yin Wong. Computer-aided cataract detection using enhanced texture features on retro-illumination lens images. In Image Processing (ICIP), 2011 18th IEEE International Conference on, pages 1565–1568. IEEE, 2011.
• Xinting Gao, Stephen Lin, and Tien Yin Wong. Automatic feature learning to grade nuclear cataracts based on deep learning. IEEE Transactions on Biomedical Engineering, 62(11):2693–2701, 2015.
• Liye Guo, Ji-Jiang Yang, Lihui Peng, Jianqiang Li, and Qingfeng Liang. A computer-aided healthcare system for cataract classification and grading based on fundus image analysis. Computers in Industry, 69:72–80, 2015.
• Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Sun Jian. Deep residual learning for image recognition. In IEEE Conference on Computer Vision & Pattern Recognition, 2016.
• Moncef Khairallah, Rim Kahloun, Rupert Bourne, Hans Limburg, Seth R Flaxman, Jost B Jonas, Jill Keeffe, Janet Leasher, Kovin Naidoo, Konrad Pesudovs, et al. Number of people blind or visually impaired by cataract worldwide and in world regions, 1990 to 2010worldwide prevalence of cataract, 1990–2010. Investigative ophthalmology & visual science, 56(11):6762–6769, 2015.
• Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. Imagenet classification with deep convolutional neural networks. In Advances in neural information processing systems, pages 1097–1105, 2012.
• Huiqi Li, Liling Ko, Joo Hwee Lim, Jiang Liu, Damon Wing Kee Wong, and Tien Yin Wong. Image based diagnosis of cortical cataract. In Engineering in Medicine and Biology Society, 2008. EMBS 2008. 30th Annual International Conference of the IEEE, pages 3904–3907. IEEE, 2008.
• Huiqi Li, Joo Hwee Lim, Jiang Liu, Paul Mitchell, Ava Grace Tan, Jie Jin Wang, and Tien Yin Wong. A computer-aided diagnosis system of nuclear cataract. IEEE Transactions on Biomedical Engineering, 57(7):1690–1698, 2010.
• Huiqi Li, Joo Hwee Lim, Jiang Liu, DamonWing KeeWong, Ngan Meng Tan, Shijian Lu, Zhuo Zhang, and Tien Yin Wong. An automatic diagnosis system of nuclear cataract using slit-lamp images. In Engineering in Medicine and Biology Society, 2009. EMBC 2009. Annual International Conference of the IEEE, pages 3693–3696. IEEE, 2009.
• Yitao Liang, Lianlian He, Chao Fan, Feng Wang, and Wei Li. Preprocessing study of retinal image based on component extraction. In IT in Medicine and Education, 2008. ITME 2008. IEEE International Symposium on, pages 670–672. IEEE, 2008.
• Hualei Shen, Hongwei Hao, Lihong Wei, and Zhibin Wang. An image based classification method for cataract. In Computer Science and Computational Technology, 2008. ISCSCT’ 08. International Symposium on, volume 1, pages 583– 586. IEEE, 2008.
• Karen Simonyan and Andrew Zisserman. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556, 2014.
• Ruchir Srivastava, Xinting Gao, Fengshou Yin, Damon WK Wong, Jiang Liu, Carol Y Cheung, and Tien YinWong. Automatic nuclear cataract grading using image gradients. Journal of Medical Imaging, 1(1):014502–014502, 2014.
• Li Xiong, Huiqi Li, and Liang Xu. An approach to evaluate blurriness in retinal images with vitreous opacity for cataract diagnosis. Journal of Healthcare Engineering, 2017, 2017.
• Ji-Jiang Yang, Jianqiang Li, Ruifang Shen, Yang Zeng, Jian He, Jing Bi, Yong Li, Qinyan Zhang, Lihui Peng, and Qing Wang. Exploiting ensemble learning for automatic cataract detection and grading. Computer methods and programs in biomedicine, 124:45–57, 2016.
• Meimei Yang, Ji-Jiang Yang, Qinyan Zhang, Yu Niu, and Jianqiang Li. Classification of retinal image for automatic cataract detection. In e-Health Networking, Applications & Services (Healthcom), 2013 IEEE 15th International Conference on, pages 674–679. IEEE, 2013.
• Matthew D Zeiler and Rob Fergus. Visualizing and understanding convolutional networks. In European conference on computer vision, pages 818–833. Springer, 2014.
• Jin Zheng, Liye Guo, Lihui Peng, Jianqiang Li, Jijiang Yang, and Qingfeng Liang. Fundus image based cataract classification. In Imaging Systems and Techniques (IST), 2014 IEEE International Conference on, pages 90–94. IEEE, 2014.