Meme Kanseri Tümörlerinin Derin Öğrenme Algoritmaları ile Sınıflandırılması

Yıl 2021, , 212 - 222, 31.12.2021

Seda Nur Özgür , Sinem Bozkurt Keser

https://doi.org/10.46810/tdfd.957618

Cited By: 2

Öz

Meme kanseri, kadınlarda ölümlere neden olabilen hastalıklar arasında en başlarda gelen hastalıklardan biridir. Yapılan araştırmalara göre meme kanserinin erken teşhisi ile ölüm oranları düşürülebilmektedir. Meme kanserinin teşhisinde incelenen mamogram görüntülerinin radyologlar tarafından incelenmesi uzun zaman almakta hatta zaman zaman bu incelemelerde hatalı sonuçlar elde edilebilmektedir. Meme kanserinin erken aşamalarda teşhis edilebilmesi için yapay zekâ yöntemleri kullanılarak yapılan çalışmalar oldukça önemlidir. Gelişen teknolojiyle birlikte birçok farklı derin öğrenme modeli bu hastalığın teşhisinde kullanılmaktadır. Bu çalışmada, meme kanserinin teşhisi için Inception-ResNet-V2 derin öğrenme modeli önerilmektedir. Önerilen derin öğrenme modeli, Inception ve ResNet modellerinin melezi bir mimari olup etkili bir şekilde geliştirilmiş sınıflandırma ve tanıma performansına sahiptir. Önerilen derin öğrenme mimarisi sırasıyla önişleme, sınıflandırma ve performans değerlendirme olmak üzere üç aşamadan oluşmaktadır. Önerilen model ile %96.21 doğruluk, %97.48 geri çağırma, %98.18 kesinlik, %97.83 F-ölçütü, %98.00 eğri altında kalan alan ve 0.83 cohen kappa performans değerleri elde edilmiştir. Elde edilen sonuçlar, çalışmada kıyaslama aşamasında kullanılan diğer derin öğrenme mimarilerinden elde edilen sonuçlar ile karşılaştırıldığında önerilen modelin meme kanseri teşhisinde daha iyi performans sergilediğini kanıtlamaktadır.

Anahtar Kelimeler

Meme Kanseri, Mamogram, Derin Öğrenme, Sınıflandırma, Inception- ResNet-V2

Classification of Breast Cancer Tumors with Deep Learning Algorithms

Öz

Breast cancer is one of the leading diseases among diseases that can cause death in women. Studies have shown that early detection of breast cancer can reduce mortality rates. The mammogram images used in the diagnosis of breast cancer are examined by radiologists. Investigations take a long time and sometimes erroneous results can be obtained. Studies using artificial intelligence methods are very important for the diagnosis of breast cancer in the early stages. It has also been observed that the diagnoses obtained with mammogram images are better than the diagnosis of medical experts. With the developing technology, many different deep learning models are used in the diagnosis of the disease. In this study, the Inception-ResNet-V2 deep learning model is proposed for the diagnosis of breast cancer. The proposed deep learning model is a hybrid architecture of Inception and ResNet models and has effectively improved classification and diagnosis performance. The proposed deep learning architecture consists of three stages, namely preprocessing, classification, and performance evaluation, respectively. With the proposed model, 96.21% accuracy, 97.48% recall, 98.18% precision, 97.83% F-score, 98.00% area under the curve, and 0.83 cohen kappa performance values were obtained. The obtained results prove that the proposed model performs better in breast cancer diagnosis when compared with the results obtained from other deep learning architectures used in the benchmarking stage in the study.

Anahtar Kelimeler

Breast Cancer, Mammogram, Deep Learning, Classification, Inception-ResNet-V2

Kaynakça

• Anonim [İnternet]. Kanser Nedir?; 2021[Erişim 5 Haziran 2021] Erişim Linki: https://hsgm.saglik.gov.tr/tr/kanser-nedir-belirtileri
• M. Akram, M. Iqbal, M. Daniyal, A. U. Khan Awareness and current knowledge of breast cancer. Biological research. 2017;50(1): 1-23.
• F. Bray, J. Ferlay, I. Soerjomataram, R. L. Siegel, L. A. Torre, A. Jemal. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. Cancer Journal for Clinicians. 2018;68(6):394-424.
• C. P. Wild, E. Weiderpass, B. W. Stewart. World Cancer Report: Cancer Research for Cancer Prevention. International Agency for Research on Cancer. Lyon, France, http://publications.iarc.fr/586. Licence: CC BY-NC-ND 3.0 IGO, 2020.
• A. Duggento, M. Aiello, C. Cavaliere, G. L. Cascella, D. Cascella, G. Conte, et all. An ad hoc random initialization deep neural network architecture for discriminating malignant breast cancer lesions in mammographic images. Contrast media & molecular imaging. 2019.
• A. Gastounioti, E. F. Conant, D. Kontos. Beyond breast density: a review on the advancing role of parenchymal texture analysis in breast cancer risk assessment. Breast cancer research. 2016;18(1):1-12.
• M. M. Jadoon, Q. Zhang, I. U. Haq, S. Butt, A. Jadoon. Three-Class Mammogram Classification Based on Descriptive CNN Features. BioMed research international. 2017.
• P. U. Hepsağ, S. A. Özel, A. Yazıcı. Using deep learning for mammography classification. International Conference on Computer Science and Engineering (UBMK). 2017;418-423. doi: 10.1109/UBMK.2017.8093429.
• A. Nahid, M. A. Mehrabi, Y. Kong. Histopathological breast cancer image classification by deep neural network techniques guided by local clustering. BioMed research international. 2018.
• Y. Zhang, C. Pan, X. Chen, F. Wang. Abnormal breast identification by nine-layer convolutional neural network with parametric rectified linear unit and rank-based stochastic pooling. Journal of Computational Science.2018;27:57-68.
• S. Sivasakthiselvan, S. Sahoo, A. Panda, R. Mishra. Image classification toward breast cancer. International Journal of Research in Engineering and Science. 2018;6(8):129-139.
• A. Behera, S. Behera, F. Das, B. Kumar. Malignant classification of mammogram images based on deep learning. International Journal of Research in Engineering and Science. 2018;6(9):35-46.
• A. H. Ahmed, M. A. M. Salem. Mammogram Based Cancer Detection Using Deep Convolutional Neural Networks. 13th International Conference on Computer Engineering and Systems (ICCES). 2018;694-699. doi: 10.1109/ICCES.2018.8639224
• D. A. Ragab, M. Sharkas, S. Marshall, J. Ren. Breast cancer detection using deep convolutional neural networks and support vector machines. PeerJ. 2019.
• H. Li, S. Zhuang, D. Li, J. Zhao, Y. Ma. Benign and malignant classification of mammogram images based on deep learning. Biomedical Signal Processing and Control. 2019;51:347-354.
• P. B. Chanda, S. K. Sarkar. Detection and Classification of Breast Cancer in Mammographic Images Using Efficient Image Segmentation Technique. Advances in Control. Signal Processing and Energy System. 2020;107-117.
• E. Trivizakis, G. S. Ioannidis, V. D. Melissianos, G. Z. Papadakis, A. Tsatsakis, D. A. Spandidos, et all. A novel deep learning architecture outperforming ‘off‑the‑shelf’ transfer learning and feature‑based methods in the automated assessment of mammographic breast density. Oncology reports. 2019;42(5):2009-2015.
• X. Yu, W. Pang, Q. Xu, M. Liang. Mammographic image classification with deep fusion learning. Scientific Reports. 2020;10(1):1-11.
• K. He, X. Zhang, S. Ren, J. Sun, Deep residual learning for image recognition. IEEE conference on computer vision and pattern recognition. 2016. ss. 770-778.