DERİN ÖĞRENME ALGORİTMASI KULLANILARAK ENDOSKOPİ GÖRÜNTÜLERİNDE POLİPLERİN ARAŞTIRILMASI

Yıl 2022, Cilt: 30 Sayı: 3, 441 - 453, 21.12.2022

Emine Cengiz Faik Yaylak Eyyüp Gülbandılar

https://doi.org/10.31796/ogummf.1122707

Öz

Makine öğrenimindeki, özellikle derin öğrenmeyle ilgili son gelişmeler, tıbbi görüntülerdeki nesneleri tanımaya ve sınıflandırmaya yardımcı olur. Bu çalışmada endoskopi görüntüleri incelenmiş, sağlıklı ve polipli hücrelerini sınıflandırılması için derin öğrenme yöntemi kullanılmıştır. Önerilen sistem için Kütahya Evliya Çelebi Eğitim ve Araştırma Hastanesi Genel Cerrahi Anabilim Dalı Endoskopi Ünitesi arşivlerinden bir veri tabanı oluşturulmuştur. Veri tabanı 54 arşiv kaydından; 93 polip ve 216 normal görüntü içermektedir. Görüntü çoğaltma için her görüntü kendi ekseni etrafında 90 derece döndürülerek toplam 1236 görüntü elde edilmiştir. Elde edilen bu verilerden rastgele seçilen verilerin 2/3'ü modelin eğitimi için kullanılırken, kalan veriler test için ayrılmıştır. Performans sonuçlarının değişkenliğini azaltmak için K-kat Çapraz Doğrulama yöntemi kullanıldı. Bu çalışmada, derin öğrenmede en iyi sınıflandırma modelini bulmak için farklı aktivasyon ve optimizasyon fonksiyonları kullanılarak 48 farklı model oluşturulmuştur. Deneysel sonuçlara göre, modellerin doğruluğunun seçilen parametrelere bağlı olduğu; %91 doğruluk oranı ile en iyi model gizli katmandaki 64 nöron, ReLU aktivasyon fonksiyonu ve RmsProp optimizasyon yöntemi ile elde edilirken, en kötü model %76 doğruluk oranı ile gizli katmandaki 32 nöron, Tanh aktivasyonu, PmsProp optimizasyon yöntemi ile elde edilmiştir. Buna göre, derin öğrenme modellerinin tasarımı sırasında farklı aktivasyon ve optimizasyon yöntemleri kullanılarak polip görüntülerinin sınıflandırma performansı optimize edilebilir.

Anahtar Kelimeler

Derin öğrenme, Aktivasyon fonksiyonu, Optimizasyon methodu, Polip, Endoskopi

INVESTIGATION OF POLYPS IN ENDOSCOPY IMAGES BY USING DEEP LEARNING ALGORITHM

Öz

Recent advances in machine learning, particularly with regard to deep learning, help to recognize and classify objects in medical images. In this study, endoscopy images were examined and deep learning method was used to classify healthy and polyp cells. For the proposed system, a database was created from the archives of General Surgery Department Endoscopy Unit in Kutahya Evliya Celebi Training and Research Hospital. The database contains 93 polyps and 216 normal images from 54 archive records. For image multiplexing, a total of 1236 images were obtained by rotating each image 90 degrees around its axis. While 2/3 of the randomly selected data from this obtained data was used for training the model, the rest of the data was reserved for testing. K-fold Cross Validation method was used to reduce the variability of performance results. In this study, 48 different models were created by using different activation and optimization functions to find the best classification model in deep learning. According to the experimental results, it was observed that accuracy of the models depends on the selected parameters; the best model with the accuracy rate of 91% was obtained with 64 neurons in the hidden layer, ReLU activation function and RmsProp optimization method whereas the worst model with the accuracy rate of 76% was obtained with 32 neurons in the hidden layer, Tanh activation and PmsProp optimization functions. Accordingly, classification performance of polyp images can be optimized by utilizing different activation and optimization methods during the design of deep learning models.

Anahtar Kelimeler

Deep learning, Activation function, Optimization method, Polyp, Endoscopy

Kaynakça

• Bengio, Y. (2008) Learning deep architectures for AI. Foundations and Trends in Machine Learning 2(1): 1– 127. Byrne, M. F., Chapados, N., Soudan, F., Oertel, C., Pérez, M. L., Kelly, R., ... & Rex, D. K. (2019). Real-time differentiation of adenomatous and hyperplastic diminutive colorectal polyps during analysis of unaltered videos of standard colonoscopy using a deep learning model. Gut, 68(1), 94-100. http://dx.doi.org/10.1136/gutjnl-2017-314547 Castelluccio, M., Poggi, G., Sansone, C. & Verdoliva, L. (2015) Land use classification in remote sensing images by convolutional neural networks, arXiv preprint arXiv:1508.00092. https://doi.org/10.48550/arXiv.1508.00092
• Cengiz, E. (2020). (Master's Thesis). Investigation of polyps in endoscopy i̇mages by using deep learning algorithm. Eskisehir Osmangazi University Graduate School of Natural and Applied Sciences, Eskisehir, Turkey. (in Turkish)
• Liu, L., Shen, C. & Van den Hengel, A. (2015) The treasure beneath convolutional layers: Cross-convolutional-layer pooling for image classification, In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, USA, pp. 4749-4757.
• Ortac, G., & Ozcan, G. (2021). Comparative study of hyperspectral image classification by multidimensional Convolutional Neural Network approaches to improve accuracy. Expert Systems with Applications, 182, 115280. https://doi.org/10.1016/j.eswa.2021.115280
• Ozawa, T., Ishihara, S., Fujishiro, M., Kumagai, Y., Shichijo, S., & Tada, T. (2020). Automated endoscopic detection and classification of colorectal polyps using convolutional neural networks. Therapeutic advances in gastroenterology, 13,1756284820910659. https://doi.org/10.1177/1756284820910659
• Pannu, H. S., Ahuja, S., Dang, N., Soni, S., & Malhi, A. K. (2020). Deep learning based image classification for intestinal hemorrhage. Multimedia Tools and Applications, 79(29), 21941-21966. https://doi:10.1109/access.2021.3061592
• Ribeiro, E., Uhl, A. & Hafner, M. (2016) Colonic polyp classification with convolutional neural networks, In: 2016 IEEE 29th International Symposium on Computer-Based Medical Systems, Belfast and Dublin, Ireland, pp.253-258.
• Ruder, S. (2016) An overview of gradient descent optimization algorithms, arXiv preprint arXiv:1609.04747. https://doi.org/10.48550/arXiv.1609.04747
• Rustam, F., Siddique, M. A., Siddiqui, H. U. R., Ullah, S., Mehmood, A., Ashraf, I. & Choi, G. S. (2021). Wireless capsule endoscopy bleeding images classification using CNN based model. IEEE Access, 9, 33675-33688.
• Sarraf, S. & Tofighi, G. (2016), 2016 IEEE Future Technologies Conference, pp. 816-820. https://doi:10.1109/FTC.2016.7821697
• Shen, D., Guoron, W., Heung-Il, S. (2017) Deep learning in medical image analysis. Annual review of Biomedical Engineering 19, 221-248. https://doi.org/10.1146/annurev-bioeng-071516-044442
• Shin, Y. & Balasingham, I. (2017) Comparison of hand-craft feature based SVM and CNN based deep learning framework for automatic polyp classification. In: 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Jeju Island, Korea, pp. 3277-3280. https://doi:10.1109/embc.2017.8037556
• Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I. & Salakhutdinov, R. (2014) Dropout: A simple way to prevent neural networks from overfitting. Journal of Machine Learning Research, 15(1), 1929- 1958.
• Suzuki, S., Zhang, X., Homma, N. et al (2016) Mass detection using deep convolutional neural network for mammographic computer-aided diagnosis, 55th Annual Conference of the Society of Instrument and Control Engineers of Japan, Tsukuba, Japan, pp. 1382-1386. doi:10.1109/sice.2016.7749265
• Tulum, G., Osman, O., Bolat, B., Dandin, Ö., Ergin, T., & Cüce, F. (2019). Colonic Polyp Classification Using Projection Image and Convolutional Neural Network. In 2019 Scientific Meeting on Electrical-Electronics & Biomedical Engineering and Computer Science (EBBT) (pp. 1-4). IEEE. https://doi:10.1109/EBBT.2019.8741701 doi:10.1109/EBBT.2019.8741701 Yang, X., Chen, S., Ding, V. et al (2016) A deep learning approach for tumor tissue image classification. IASTED Biomedical Engineering – 2016, https://www.researchgate.net/publication/298929528_A_Deep_Learning_Approach_for_Tumor_Tissue_Image_Classification
• Yazan, E. & Talu, M.F. (2017) Comparison of the stochastic gradient descent based optimization techniques, 2017 International Artificial Intelligence and Data Processing Symposium (IDAP), Malatya, Turkey, pp. 1-5. https://doi:10.1109/IDAP.2017.8090299
• Yixuan, Y. & Meng, M. (2017) Deep learning for polyp recognition in wireless endoscopy images. Medical Physics, 44(4), 1379-1389. https://doi.org/10.1002/mp.12147
• Zou, Y., Li, L., Wang, Y. et al (2015) Classifying digestive organs in wireless capsule endoscopy images based on deep convolutional neural network. In: 2015 IEEE International Conference on Digital Signal Processing, Singapore, pp. 1274-1278. https://doi:10.1109/ICDSP.2015.7252086