Research Article
BibTex RIS Cite

Derin Öğrenme Mimarilerini Kullanarak Katarakt Tespiti

Year 2021, Issue: 28, 1428 - 1433, 30.11.2021
https://doi.org/10.31590/ejosat.1012694

Abstract

İnsanın yaşam kalitesini olumsuz olarak etkileyen görme kayıplarını daha erken bir dönemde teşhis etmek önemlidir. İnsan yaşının ilerlemesi ile birlikte görme bozuklukları ve bazen tamamen görme kaybına neden olmaktadır. Gözün anatomik yapısında bulunan anormallikler göz hastalıklarının erken dönemlerinde göz yapısına ait görsellerle de tespit edilebilmektedir. Katarat dünyada milyonlarca insanı etkileyen görme bozukluğunun en önemli nedenidir. Otomatik tanı sistemleri ile sağlık hizmeti kullanımı hafifleyerek uzmanlara yardımcı olmayı amaçlamaktadır. Bu makalede renkli fundus görüntüler kullanılarak katarat hastalığına otomatik tanı sistemi ele alınmıştır. Katarat hastalığının otomatik tanımlanması için evrişimli sinir ağı (CNN) ve derin artık ağ (DRN) kullanılarak sınıflandırma yöntemi kullanılmıştır. Veri seti 5000 hastanın sağ ve sol gözlerine ait renkli fundus fotoğrafları ve doktorların her bir hastanın sağ ve sol gözüne konulmuş teşhisler için anahtar kelimler ile yapılandırılmış bir veri tabanıdır. Bu veri seti gerçek yaşamda hasta gruplarını temsil etmektedir. Çinli bir şirket olan Shanggong Medical Technology Co., Ltd. Şirketi tarafından farklı hastane ve tıp merkezlerinden elde edilen veriler toplanmıştır. Veri setinde hastalar 8 farklı etikete sınıflandırma yapılmıştır. Renkli fundus görüntüler sayesinde farklı evrelere ait katarat semptomlarına ait özellikler bulunmaktadır. Önerilen otomatik tanı sistemi güncel sınıflandırma sistemlerine oranla daha başarılı olduğu görülmektedir. DRN yönteminin CNN yöntemine göre doğruluk oranına göre daha yüksektir. CNN modelinde doğruluk oranı %89 civarında iken DRN modelinde doğruluk oranı %95 olduğu görülmektedir.

References

  • Yang J J, Li J, Shen, R, Zeng Y, He J et al.(2016). Exploiting ensemble learning for automatic cataract detection and grading. Computer Methods and Programs in Biomedicine; 124: 45–57. doi:10.1016/j.cmpb.2015.10.007
  • Yang M, Yang J J, Zhang Q, Niu Y, Li J. Classification of retinal image for automatic cataract detection, In: IEEE International Conference on e-Health Networking, Applications Services; Lisbon, Portugal; 2013. pp. 674–679. doi:10.1109/HealthCom.2013.6720761
  • Wang Liming, Zhang K, Liu X, Long E, Jiang J, An Y et al. (2017). Comparative analysis of image classification methods 5 for automatic diagnosis of ophthalmic images. Scientific Reports; 7: 1–11. doi:10.1038/srep41545
  • Gali H E, Sella R, Afshari N A. ( 2019). Cataract grading systems: a review of past and present. Current opinion in ophthalmology; 30(1): 13-18. doi: 10.1097/ICU.0000000000000542
  • Grewal P S, Oloumi F, Rubin U, Tennant M T S. (2018). Deep learning in ophthalmology: a review. In Canadian Journal of Ophthalmology; 53(4): 309–313.doi:10.1016/j.jcjo.2018.04.019
  • He J, Li C, Ye J, Qiao Y, Gu L. (2021). Multi-label ocular disease classification with a dense correlation deep neural network. Biomedical Signal Processing and Control ; 63: doi:10.1016/j.bspc.2020.102167
  • Yoo T K, Ryu I H, Kim J K, Lee I S, Kim J S et al. (2020). Deep learning can generate traditional retinal fundus photographs using ultra-widefield images via generative adversarial networks. Computer Methods and Programs in Biomedicine; 197: doi:10.1016/j.cmpb.2020.105761
  • Long E, Lin H, Liu Z, Wu X, Wang L, et al. (2017). An artificial intelligence platform for the multihospital collaborative management of congenital cataracts. Nature biomedical engineering ; 1(2): 1-8. doi: doi.org/10.1038/s41551-016-0024
  • Zhang H, He Z. (2019). Automatic cataract grading methods based on deep learning. Computer Methods and Programs in Biomedicine; 182: doi:10.1016/j.cmpb.2019.07.006
  • Yanagihara R T, Lee C S, Ting D S W, Lee A Y. (2020). Methodological challenges of deep learning in optical coherence tomography for retinal diseases: a review. Translational Vision Science & Technology; 9(2): 11-11. doi: 10.1167/tvst.9.2.11
  • Y. Lecun, L. Bottou, Y. Bengio, and P. Haffner, (Nov. 1998). “Gradient-based learning applied to document recognition,” Proc. IEEE, vol. 86, no. 11, pp. 2278– 2324,
  • He, K., Zhang, X., Ren, S., Sun, J. (2016). Deep Residual Learning for Image Recognition. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 770– 778.
  • Liefers B, Venhuizen F G, Schreur V, van Ginneken B, Hoyng C et al. (2017). Automatic detection of the foveal center in optical coherence tomography. Biomedical Optics Express; 8(11): 5160. doi:10.1364/boe.8.005160
  • Xu X, Zhang L, Li J, Guan Y, Zhang L. (2020). A Hybrid Global-Local Representation CNN Model for Automatic Cataract Grading. IEEE Journal of Biomedical and Health Informatics; 24(2): 556–567. doi:10.1109/JBHI.2019.2914690
  • Gour N, Khanna P. (2020). Multi-class multi-label ophthalmological disease detection using transfer learning based convolutional neural network. Biomedical Signal Processing and Control ; doi: doi.org/10.1016/j.bspc.2020.102329
  • Pratap T, Kokil P. (2019). Computer-aided diagnosis of cataract using deep transfer learning. Biomedical Signal Processing and Control; 53: 1-8. doi: 10.1016/j.bspc.2019.04.010
  • Sengupta S, Singh A, Leopold H A, Gulati T, Lakshminarayanan, V. (2020). Ophthalmic diagnosis using deep learning with fundus images–A critical review. Artificial intelligence in medicine ; 102: doi: 10.1016/j.artmed.2019.101758

Cataract Detection Using Deep Learning Architectures

Year 2021, Issue: 28, 1428 - 1433, 30.11.2021
https://doi.org/10.31590/ejosat.1012694

Abstract

It is important to diagnose vision loss, which negatively affects the quality of life of people, at an earlier stage. With the advancement of human age, it causes visual disturbances and sometimes complete vision loss. Abnormalities in the anatomical structure of the eye can also be detected with visuals of the eye structure in the early stages of eye diseases. Cataracts are the most important cause of visual impairment affecting millions of people around the world. It aims to help experts by reducing the use of health services with automatic diagnosis systems. In this article, the automatic diagnosis system for catarrhal disease using color fundus images is discussed. Classification method using convolutional neural network (CNN) and deep residual network (DRN) was used for automatic identification of cataract disease. The dataset is a structured database with color fundus photographs of 5000 patients' right and left eyes and keywords for doctors to diagnose each patient's right and left eyes. This dataset represents real-life patient groups. Shanggong Medical Technology Co., Ltd., a Chinese company. Data from different hospitals and medical centers were collected by the company. In the data set, patients were classified into 8 different labels. Thanks to the color fundus images, there are features of cataract symptoms of different stages. It is seen that the proposed automatic diagnosis system is more successful than the current classification systems. The accuracy rate of the DRN method is higher than the CNN method. While the accuracy rate in the CNN model is around 89%, the accuracy rate in the DNN model is 95%.

References

  • Yang J J, Li J, Shen, R, Zeng Y, He J et al.(2016). Exploiting ensemble learning for automatic cataract detection and grading. Computer Methods and Programs in Biomedicine; 124: 45–57. doi:10.1016/j.cmpb.2015.10.007
  • Yang M, Yang J J, Zhang Q, Niu Y, Li J. Classification of retinal image for automatic cataract detection, In: IEEE International Conference on e-Health Networking, Applications Services; Lisbon, Portugal; 2013. pp. 674–679. doi:10.1109/HealthCom.2013.6720761
  • Wang Liming, Zhang K, Liu X, Long E, Jiang J, An Y et al. (2017). Comparative analysis of image classification methods 5 for automatic diagnosis of ophthalmic images. Scientific Reports; 7: 1–11. doi:10.1038/srep41545
  • Gali H E, Sella R, Afshari N A. ( 2019). Cataract grading systems: a review of past and present. Current opinion in ophthalmology; 30(1): 13-18. doi: 10.1097/ICU.0000000000000542
  • Grewal P S, Oloumi F, Rubin U, Tennant M T S. (2018). Deep learning in ophthalmology: a review. In Canadian Journal of Ophthalmology; 53(4): 309–313.doi:10.1016/j.jcjo.2018.04.019
  • He J, Li C, Ye J, Qiao Y, Gu L. (2021). Multi-label ocular disease classification with a dense correlation deep neural network. Biomedical Signal Processing and Control ; 63: doi:10.1016/j.bspc.2020.102167
  • Yoo T K, Ryu I H, Kim J K, Lee I S, Kim J S et al. (2020). Deep learning can generate traditional retinal fundus photographs using ultra-widefield images via generative adversarial networks. Computer Methods and Programs in Biomedicine; 197: doi:10.1016/j.cmpb.2020.105761
  • Long E, Lin H, Liu Z, Wu X, Wang L, et al. (2017). An artificial intelligence platform for the multihospital collaborative management of congenital cataracts. Nature biomedical engineering ; 1(2): 1-8. doi: doi.org/10.1038/s41551-016-0024
  • Zhang H, He Z. (2019). Automatic cataract grading methods based on deep learning. Computer Methods and Programs in Biomedicine; 182: doi:10.1016/j.cmpb.2019.07.006
  • Yanagihara R T, Lee C S, Ting D S W, Lee A Y. (2020). Methodological challenges of deep learning in optical coherence tomography for retinal diseases: a review. Translational Vision Science & Technology; 9(2): 11-11. doi: 10.1167/tvst.9.2.11
  • Y. Lecun, L. Bottou, Y. Bengio, and P. Haffner, (Nov. 1998). “Gradient-based learning applied to document recognition,” Proc. IEEE, vol. 86, no. 11, pp. 2278– 2324,
  • He, K., Zhang, X., Ren, S., Sun, J. (2016). Deep Residual Learning for Image Recognition. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 770– 778.
  • Liefers B, Venhuizen F G, Schreur V, van Ginneken B, Hoyng C et al. (2017). Automatic detection of the foveal center in optical coherence tomography. Biomedical Optics Express; 8(11): 5160. doi:10.1364/boe.8.005160
  • Xu X, Zhang L, Li J, Guan Y, Zhang L. (2020). A Hybrid Global-Local Representation CNN Model for Automatic Cataract Grading. IEEE Journal of Biomedical and Health Informatics; 24(2): 556–567. doi:10.1109/JBHI.2019.2914690
  • Gour N, Khanna P. (2020). Multi-class multi-label ophthalmological disease detection using transfer learning based convolutional neural network. Biomedical Signal Processing and Control ; doi: doi.org/10.1016/j.bspc.2020.102329
  • Pratap T, Kokil P. (2019). Computer-aided diagnosis of cataract using deep transfer learning. Biomedical Signal Processing and Control; 53: 1-8. doi: 10.1016/j.bspc.2019.04.010
  • Sengupta S, Singh A, Leopold H A, Gulati T, Lakshminarayanan, V. (2020). Ophthalmic diagnosis using deep learning with fundus images–A critical review. Artificial intelligence in medicine ; 102: doi: 10.1016/j.artmed.2019.101758
There are 17 citations in total.

Details

Primary Language Turkish
Subjects Engineering
Journal Section Articles
Authors

Fatih Ağalday 0000-0002-2635-0661

Ahmet Çınar 0000-0003-4324-4964

Publication Date November 30, 2021
Published in Issue Year 2021 Issue: 28

Cite

APA Ağalday, F., & Çınar, A. (2021). Derin Öğrenme Mimarilerini Kullanarak Katarakt Tespiti. Avrupa Bilim Ve Teknoloji Dergisi(28), 1428-1433. https://doi.org/10.31590/ejosat.1012694