sapiСистемный анализ и прикладная информатика«System analysis and applied information science»2309-49232414-0481Belarusian National Technical University10.21122/2309-4923-2022-3-12-21sapi-577Research ArticleСистемный анализSystem analysisКлассификация стадий диабетической ретинопатии на основе нейронных сетейClassification of diabetic retinopathy stages based on neural networksЛукашевичМ. М.LukashevichM. M.

Лукашевич Марина Михайловна, кандидат технических наук, доцент, докторант

Минск

PhD, Associate Professor, Postdoctoral Fellow

Minsk

lukashevich@bsuir.byГолубЮ. И.GolubY. I.

Голуб Юлия Игоревна, кандидат технических наук, доцент, старший научный сотрудник 

Минск

PhD, Associate Professor, Senior Research Fellow

Minsk

6423506@gmail.comБелорусский государственный университет информатики и радиоэлектроникиБеларусьBelarusian State University of Informatics and RadioelectronicsBelarusОбъединенный институт проблем информатики Национальной академии наук БеларусиБеларусьUnited Institute of Informatics Problems, National Academy of Sciences of BelarusBelarus202205122022031221Copyright © Лукашевич М.М., Голуб Ю.И., 20222022Лукашевич М.М., Голуб Ю.И.Lukashevich M.M., Golub Y.I.Данная работа распространяется под лицензией Creative Commons Attribution 4.0.This work is licensed under a Creative Commons Attribution 4.0 License.https://sapi.bntu.by/jour/article/view/577

Диабетическая ретинопатия является одним из основных побочных эффектов диабета, которая вызывает тяжелые последствия, вплоть до слепоты. Основной задачей является ранняя диагностика данного заболевания с целью своевременного и эффективного лечения. Диабетическая ретинопатия может быть обнаружена гораздо быстрее и более точно, если использовать методы машинного обучения для анализа изображений сетчатки глаза человека. Разработка методов и алгоритмов детекции и классификации данного заболевания, а также автоматизация этого процесса, являются актуальными задачами и экономически эффективным мероприятием.

В статье основное внимание уделено вопросу классификации стадий диабетической ретинопатии с помощью нейронных сетей на основе изображений сетчатки глаза человека. Oписана задача классификации стадий диабетической ретинопатии, а также предложено использование архитектуры глубоких нейронных сетей на основе VGG16 и VGG19 с добавление пользовательских слоёв. В результате проведенных экспериментальных исследований приведены рекомендации по выбору размера исходных изображений сетчатки глаза, а также этапу предварительной обработки (обрезке изображения).

Выполнено исследование используемого набора данных, в результате чего обучение моделей нейронных сетей и оценка результатов проводилась с учетом дисбаланса классов.

Diabetic retinopathy is one of the main side effects of diabetes, which causes severe effects, including blindness. The main challenge is the early diagnosis of this disease for timely and effective treatment. Diabetic retinopathy can be detected much faster and more accurately by using machine learning methods for image analyzing of the human retina. The development of methods and algorithms for the detection and classification of this disease, the automation of this process are the actual and costeffective goals.

The article focuses on the classification of the stages of diabetic retinopathy using neural networks based on human retinal images. Classification problem of diabetic retinopathy stages is described.

The architecture of deep neural networks based on VGG16 and VGG19 with the addition of custom layers is proposed. Recommendations for the selection of the size of the initial retinal images and the preprocessing stage (cropping) are given As a result of the performed experimental research. Analysis of the dataset was performed. Neural network models were trained and results were evaluated with class imbalance taken into account.

изображениеклассификациядиабетическая ретинопатиянейронные сетиimageclassificationdiabetic retinopathyneural networksReferencesBourne, R.R. Vision Loss Expert Group. Causes of vision loss worldwide, 1990-2010: a systematic analysis / R.R. Bourne, G.A. Stevens, R.A. White, J.L. Smith, S.R. Flaxman, H. Price, J.B. Jonas, J. Keeffe, J. Leasher, K. Naidoo, K. Pesudovs, S. Resnikoff, H.R. Taylor // Lancet Glob Health 2013. – T. 1, №6. C. 339-349. doi: 10.1016/S2214-109X(13)70113-X.Bourne, R.R. Vision Loss Expert Group. Causes of vision loss worldwide, 1990-2010: a systematic analysis / R.R. Bourne, G.A. Stevens, R.A. White, J.L. Smith, S.R. Flaxman, H. Price, J.B. Jonas, J. Keeffe, J. Leasher, K. Naidoo, K. Pesudovs, S. Resnikoff, H.R. Taylor // Lancet Glob Health 2013. – Vol. 1, №6. P. 339-349. doi: 10.1016/S2214-109X(13)70113-X.Diabetic Retinopathy: National Eye Institute [Электронный ресурс]. – Режим доступа: https://nei.nih.gov/health/diabetic/retinopathy. – Дата доступа : 17.08.2022.Diabetic Retinopathy: National Eye Institute [Online]. – Available: https://nei.nih.gov/health/diabetic/retinopathy. – Date of access : 17.08.2022.World Health Organization: Diabetes [Электронный ресурс]. – Режим доступа: https://www .who.int/news-room/ fact-sheets/detail/diabetes. – Дата доступа : 17.08.2022.World Health Organization: Diabetes [Online]. – Available: https://www .who.int/news-room/fact-sheets/detail/diabetes. – Date of access : 17.08.2022.Stolte, S. A survey on medical image analysis in diabetic retinopathy / S. Stolte, R. Fang // Medical Image Analysis. – 2020. – T. 64. – C. 1-76. https://doi.org/10.1016/j.media.2020.101742.Stolte, S. A survey on medical image analysis in diabetic retinopathy / S. Stolte, R. Fang // Medical Image Analysis. – 2020. – Vol. 64. – P. 1-76. https://doi.org/10.1016/j.media.2020.101742.Gulshan, V. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs / V. Gulshan [et al.] // JAMA. – 2016. – T. 316, № 2. – С. 2402-2410 (2016). https://doi.org/10.1001/jama.2016.17216.Gulshan, V. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs / V. Gulshan [et al.] // JAMA. – 2016. – Vol. 316, № 2. – P. 2402-2410 (2016). https://doi.org/10.1001/jama.2016.17216.Gargeya, R. Automated identifi of diabetic retinopathy using deep learning / R. Gargeya, T. Leng // Ophthalmology. – 2017. – T. 124, № 7. – C. 962–969. doi: 10.1016/j.ophtha.2017.02.008.Gargeya, R. Automated identifi of diabetic retinopathy using deep learning / R. Gargeya, T. Leng // Ophthalmology. – 2017. – Vol. 124, № 7. – P. 962–969. doi: 10.1016/j.ophtha.2017.02.008.Alyoubi, W.L. Diabetic Retinopathy Detection through Deep Learning Technique: A Review / W.L. Alyoubi, W.M. Shalash, M.F. Abulkhair // Informatics in Medicine Unlocked. 2020. T. 20. – C. 1–11. doi: 10.1016/j.imu.2020.100377.Alyoubi, W. L. Diabetic Retinopathy Detection through Deep Learning Technique: A Review / W.L. Alyoubi, W.M. Shalash, M.F. Abulkhair // Informatics in Medicine Unlocked. 2020. Vol. 20. – P. 1–11. doi: 10.1016/j.imu.2020.100377.Li, T. Diagnostic assessment of deep learning algorithms for diabetic retinopathy screening / T. Li [et al.] // Information Sciences. 2019. T. 501.C. 511–522. https://doi.org/10.1016/j.ins.2019.06.011Li, T. Diagnostic assessment of deep learning algorithms for diabetic retinopathy screening / T. Li [et al.] // Information Sciences. 2019. Vol. 501. P. 511–522. https://doi.org/10.1016/j.ins.2019.06.011APTOS 2019 Blindness Detection [Электронный ресурс]. – Режим доступа: https://www.kaggle.com/competitions/aptos2019-blindness-detection/data. – Дата доступа : 17.08.2022.APTOS 2019 Blindness Detection [Online]. – Available: https://www.kaggle.com/competitions/aptos2019-blindness-detection/data. – Date of access : 17.08.2022.Sim, D.A. Patterns of peripheral retinal and central macula ischemia in diabetic retinopathy as evaluated by ultra-widefi fl angiography / D.A. Sim [et al.] // American Journal of Ophthalmology. –2014. – T. 158. – C. 144–153.Sim, D. A. Patterns of peripheral retinal and central macula ischemia in diabetic retinopathy as evaluated by ultra-widefi fl angiography / D.A. Sim [et al.] // American Journal of Ophthalmology. –2014. – Vol. 158. – P. 144–153.Wejdan L. Alyoubi Diabetic retinopathy detection through deep learning techniques: A review / Wejdan L. Alyoubi, Wafaa M. Shalash, Maysoon F. Abulkhair // Informatics in Medicine Unlocked. – 2020. – T. 20. – C. 1-11. https://doi.org/10.1016/j.imu.2020.100377.Wejdan L. Alyoubi Diabetic retinopathy detection through deep learning techniques: A review / Wejdan L. Alyoubi, Wafaa M. Shalash, Maysoon F. Abulkhair // Informatics in Medicine Unlocked. – 2020. – Vol. 20. – P. 1-11. https://doi.org/10.1016/j.imu.2020.100377.Simonyan K., Zisserman A. Very deep convolutional networks for large-scale image recognition [Электронный ресурс]. – Режим доступа: https://arxiv.org/pdf/1409.1556. – Дата доступа : 17.08.2022.Simonyan K., Zisserman A. Very deep convolutional networks for large-scale image recognition [Online]. – Available: https://arxiv.org/pdf/1409.1556. – Date of access : 17.08.2022.Géron, A. Hands-on machine learning with Scikit-Learn, Keras and TensorFlow: concepts, tools, and techniques to build intelligent systems (2nd ed.) / A. Géron O’Reilly Media, Inc., 2019. – 856 P.Géron, A. Hands-on machine learning with Scikit-Learn, Keras and TensorFlow: concepts, tools, and techniques to build intelligent systems (2nd ed.) / A. Géron O’Reilly Media, Inc., 2019. – 856 P.Zhuang, F. A comprehensive survey on transfer learning / F. Zhuang [et al.] // Proceedings of the IEEE. – 2020. – Т. 109. – №. 1. – С. 43-76.Zhuang, F. A comprehensive survey on transfer learning / F. Zhuang [et al.] // Proceedings of the IEEE. – 2020. – Vol. 109. – №. 1. – P. 43-76.Russakovsky, O. Imagenet large scale visual recognition challenge / Olga Russakovsky, Jia Deng, Hao Su, Jonathan Krause, Sanjeev Satheesh, Sean Ma, Zhiheng Huang, Andrej Karpathy, Aditya Khosla, Michael Bernstein, et al. // International Journal of Computer Vision. – 2015. – T. 115, № 3. – C. 211–252.Russakovsky, O. Imagenet large scale visual recognition challenge / Olga Russakovsky, Jia Deng, Hao Su, Jonathan Krause, Sanjeev Satheesh, Sean Ma, Zhiheng Huang, Andrej Karpathy, Aditya Khosla, Michael Bernstein, et al. // International Journal of Computer Vision. – 2015. – Vol. 115, № 3. – P. 211–252.Srivastava, N. Dropout: a simple way to prevent neural networks from overfi / Nitish Srivastava, Geoffrey Hinton, Alex Krizhevsky, Ilya Sutskever, Ruslan Salakhutdinov // Journal of Machine Learning Research – 2014 T. 15, № 1. C. 1929–1958.Srivastava, N. Dropout: a simple way to prevent neural networks from overfi / Nitish Srivastava, Geoffrey Hinton, Alex Krizhevsky, Ilya Sutskever, Ruslan Salakhutdinov // Journal of Machine Learning Research – 2014 Vol. 15, № 1. P. 1929–1958.Ioffe S. Batch normalization: Accelerating deep network training by reducing internal covariate shift / S. Ioffe, C. Szegedy // International conference on machine learning. – PMLR. 2015. – С. 448-456.Ioffe S. Batch normalization: Accelerating deep network training by reducing internal covariate shift / S. Ioffe, C. Szegedy // International conference on machine learning. – PMLR. 2015. – P. 448-456.Brock A. et al. Freezeout: Accelerate training by progressively freezing layers [Электронный ресурс]. – Режим доступа: https://arxiv.org/pdf/1706.04983. – Дата доступа : 17.08.2022.Brock A. et al. Freezeout: Accelerate training by progressively freezing layers [Online]. – Available: https://arxiv.org/pdf/1706.04983. – Date of access : 17.08.2022.Ruder S. An overview of gradient descent optimization algorithms [Электронный ресурс]. – Режим доступа: https://arxiv.org/pdf/1609.04747. – Дата доступа : 17.08.2022.Ruder S. An overview of gradient descent optimization algorithms [Online]. – Available: https://arxiv.org/ pdf/1609.04747. – Date of access : 17.08.2022.Goutte, C. A Probabilistic Interpretation of Precision, Recall and F-Score, with Implication for Evaluation / Goutte, Cyril, Gaussier, Eric.// Lecture Notes in Computer Science. – 2005. – T. 3408. – C.345-359. doi: 10.1007/978-3-540-31865-1_25.Goutte, C. A Probabilistic Interpretation of Precision, Recall and F-Score, with Implication for Evaluation / Goutte, Cyril, Gaussier, Eric.// Lecture Notes in Computer Science. – 2005. – Vol. 3408. – P. 345-359. doi:10.1007/978-3-540-31865-1_25.Salamat, N. Diabetic retinopathy techniques in retinal images: A review / N. Salamat, MMS. Missen, A. Rashid // Artifi Intelligence in Medicine. – 2019. – T. 97. – C. 168-188. doi: 10.1016/j.artmed.2018.10.009.Salamat, N. Diabetic retinopathy techniques in retinal images: A review / N. Salamat, MMS. Missen, A. Rashid // Artifi Intelligence in Medicine. – 2019. – Vol. 97. – P. 168-188. doi: 10.1016/j.artmed.2018.10.009.Y.I. Starovoitov, V.V. A Universal Retinal Image Template for Automated Screening of Diabetic Retinopathy / V.V. Starovoitov, Golub, M.M. Lukashevich // Pattern Recognition and Image Analysis. – 2022. – T. 32. – C. 322–331. https://doi.org/10.1134/S1054661822020195.Starovoitov, V. V. A Universal Retinal Image Template for Automated Screening of Diabetic Retinopathy / V.V. Starovoitov, Y.I. Golub, M.M. Lukashevich // Pattern Recognition and Image Analysis. – 2022. – Vol. 32. – P. 322–331. https://doi.org/10.1134/S1054661822020195.

The authors declare that there are no conflicts of interest present.