Cancer Detection and Location Methods in Histopathological Images

Öz

Tumor detection in the histopathological images of breast lymph nodes is one of the most important findings in the diagnosis of breast cancer. Histopathological images are carefully examined by pathologists and tumor detection is performed. This process causes both workload density and a subjective assessment. The CAMELYON16 dataset has been used by the International Symposium on Biomedical Image (ISBI) for automatic detection of the tumor on histopathological images. Data sets created at different level resolutions of whole slide images (WSI) have been analyzed using Faster RCNN. The data sets with different image sizes created from 3rd resolution levels of WSI have been analyzed using Mask RCNN. Finally, the HMS & MIT method, which ranked in ISBI, has been applied on a limited data set, and its performance on the histopathological image data sets has been compared with the Faster RCNN and Mask RCNN algorithms. Although Mask RCNN (57% AUC) has been trained with low-resolution level images (3rd levels) has an accuracy value close to the HMS & MIT (58% AUC) (using high-resolution level images, 0th level) method. Also, for a WSI analysis, the result has been obtained with Faster RCNN as soon as possible (0.58 hours)

Anahtar Kelimeler

• Semantic Segmentation
• Deep Learning
• Faster RCNN
• Medical Image Processing
• Mask RCNN
• Object Detection
• Tumor Detection

Histopatolojik Görüntülerde Kanser Tespit ve Lokasyon Yöntemleri

Öz

Meme lenf düğümlerinin histopatolojik görüntülerinde tümör tespiti meme kanseri teşhisinde en önemli bulgulardan bir tanesidir. Histopatolojik görüntüler, patologlar tarafından dikkatli bir şekilde incelenerek tümör tespiti yapılır. Bu işlem hem iş yükü yoğunluğuna hem de sübjektif bir değerlendirmeye neden olmaktadır. Görüntülerde tümörün otomatik tespiti için International Symposium on Biomedical Image (ISBI) tarafından Camelyon16 veri seti oluşturulmuştur. Bu veri seti, hızlı bölge tabanlı evrişimli sinir ağı (Faster Region-Based Convolutional Neural Network, Faster RCNN) ve mask bölge tabanlı evrişimli sinir ağı (Mask Region-Based Convolutional Neural Network, Mask RCNN) derin öğrenme algoritmaları kullanılarak tümör tespiti ve bölütleme yapılmıştır. Tüm slayt görüntülerin farklı seviye çözünürlüklerinde oluşturulan veri setleri ile Faster RCNN ve görüntülerin 3. çözünürlük seviyeden oluşturulan farklı boyutu veri setleri ile Mask RCNN algoritmaları performansları incelenmiştir. Son olarak ISBI’da dereceye giren HMS&MIT yöntemi kısıtlı bir veri seti üzerinde uygulanarak Faster RCNN ve Mask RCNN algoritmaları ile tüm slayt görüntüsü üzerindeki başarımları kıyaslanmıştır. Tüm slayt görüntülerinin analizinde Mask RCNN (%57) görüntülerin düşük çözünürlük seviyesinde (3. seviye) çalışmış olmasına rağmen yüksek çözünürlük seviyesinde (0. seviye) çalışan HMS&MIT (%58) yöntemine yakın bir doğruluk (AUC) değeri almaktadır.

Anahtar Kelimeler

• Görüntü Bölütleme
• Derin Ögrenme
• Faster RCNN
• Mask RCNN
• Medikal Görüntü Bölütleme
• Nesne Tespiti
• Tümör Tespiti

Kaynakça

1. American Cancer Society. (2019). Surveillance Research, 5. Retrieved from https://www.cancer.org/content/dam/cancer-org/research/cancer-facts-and-statistics/annual-cancer-facts-and-figures/2019/cancer-facts-and-figures-2019.pdf
2. Babak Ehteshami, B., Geert, L., Nadya, T., Irene, O.-H., André, H., Nico, K., & Jeroen A W M, V. D. L. (2016). Stain specific standardization of whole-slide histopathological images. IEEE Transactions on Medical Imaging, 35(2), 404–415. Retrieved from https://ieeexplore.ieee.org/abstract/document/7243333/
3. Bejnordi, B. E., Litjens, G., Hermsen, M., Karssemeijer, N., & van der Laak, J. A. W. M. (2015). A multi-scale superpixel classification approach to the detection of regions of interest in whole-slide histopathology images. Medical Imaging 2015: Digital Pathology, 9420, 94200H.
4. Celik, Y., Talo, M., Yildirim, O., Karabatak, M., & Acharya, U. R. (2020). Automated Invasive Ductal Carcinoma Detection Based Using Deep Transfer Learning with Whole-Slide Images. Pattern Recognition Letters. Elsevier B.V. Retrieved from https://doi.org/10.1016/j.patrec.2020.03.011
5. Chollet, F. (2017). Xception: Deep Learning with Depthwise Separable Convolutions. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 1800–1807). IEEE. Retrieved from http://ieeexplore.ieee.org/document/8099678/
6. Ehteshami Bejnordi, B., Balkenhol, M., Litjens, G., Holland, R., Bult, P., Karssemeijer, N., & Van Der Laak, J. A. W. M. (2016). Automated Detection of DCIS in Whole-Slide H&E Stained Breast Histopathology Images. IEEE Transactions on Medical Imaging, 35(9), 2141–2150.
7. Ehteshami Bejnordi, B., Veta, M., Johannes van Diest, P., van Ginneken, B., Karssemeijer, N., Litjens, G., van der Laak, J. A. W. M., et al. (2017). Diagnostic Assessment of Deep Learning Algorithms for Detection of Lymph Node Metastases in Women With Breast Cancer. JAMA, 318(22), 2199. Retrieved from http://jama.jamanetwork.com/article.aspx?doi=10.1001/jama.2017.14585
8. Fan, J., Upadhye, S., & Worster, A. (2006). Understanding receiver operating characteristic (ROC) curves. CJEM, 8(01), 19–20. Retrieved from https://www.cambridge.org/core/product/identifier/S1481803500013336/type/journal_article

9. Girshick, R. (2015). Fast R-CNN. Proceedings of the IEEE International Conference on Computer Vision, 2015 Inter, 1440–1448.
10. Girshick, R., Donahue, J., Darrell, T., & Malik, J. (2016). Region-Based Convolutional Networks for Accurate Object Detection and Segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 38(1), 142–158. Retrieved from http://ieeexplore.ieee.org/document/7112511/
11. Guo, Z., Liu, H., Ni, H., Wang, X., Su, M., Guo, W., Wang, K., et al. (2019). A Fast and Refined Cancer Regions Segmentation Framework in Whole-slide Breast Pathological Images. Scientific Reports, 9(1), 1–10. Springer US. Retrieved from http://dx.doi.org/10.1038/s41598-018-37492-9
12. Gupta, V., & Bhavsar, A. (2018). Sequential modeling of deep features for breast cancer histopathological image classification. IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, 2018-June, 2335–2342.
13. He, K., Gkioxari, G., Dollár, P., & Girshick, R. (2017). Mask R-CNN. 2017 IEEE International Conference on Computer Vision (ICCV), 2017-Octob, 2980–2988. IEEE. Retrieved from http://ieeexplore.ieee.org/document/8237584/
14. Ishikawa, M., Okamoto, C., Shinoda, K., Komagata, H., Iwamoto, C., Ohuchida, K., Hashizume, M., et al. (2019). Detection of pancreatic tumor cell nuclei via a hyperspectral analysis of pathological slides based on stain spectra. Biomedical Optics Express, 10(9), 4568. Retrieved from https://www.osapublishing.org/abstract.cfm?URI=boe-10-9-4568
15. Reis, S., Gazinska, P., Hipwell, J. H., Mertzanidou, T., Naidoo, K., Williams, N., Pinder, S., et al. (2017). Automated Classification of Breast Cancer Stroma Maturity from Histological Images. IEEE Transactions on Biomedical Engineering, 64(10), 2344–2352.
16. Ren, S., He, K., Girshick, R., & Sun, J. (2015). Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. Computer Vision and Pattern Recognition, 1–14. Retrieved from https://arxiv.org/abs/1506.01497
17. Riaz, N., Wolden, S. L., Gelblum, D. Y., & Eric, J. (2016). Multi-instance Multi-label Learning for Multi-class Classification of Whole Slide Breast Histopathology Images, 118(24), 6072–6078.
18. Samah, A. A., Fauzi, M. F. A., & Mansor, S. (2017). Classification of benign and malignant tumors in histopathology images. Proceedings of the 2017 IEEE International Conference on Signal and Image Processing Applications, ICSIPA 2017, 102–106.
19. Song, Y., Zou, J. J., Chang, H., & Cai, W. (2017). Adapting fisher vectors for histopathology image classification. Proceedings - International Symposium on Biomedical Imaging, 600–603. IEEE.
20. Spanhol, F. A., Oliveira, L. S., Petitjean, C., & Heutte, L. (2016). A Dataset for Breast Cancer Histopathological Image Classification. IEEE Transactions on Biomedical Engineering, 63(7), 1455–1462.
21. Verma, R., Sharma, S., Vahadane, A., Kumar, N., Sethi, A., & Bhargava, S. (2017). A Dataset and a Technique for Generalized Nuclear Segmentation for Computational Pathology. IEEE Transactions on Medical Imaging, 36(7), 1550–1560.
22. Wan, S., Lee, H. C., Huang, X., Xu, T., Xu, T., Zeng, X., Zhang, Z., et al. (2017). Integrated local binary pattern texture features for classification of breast tissue imaged by optical coherence microscopy. Medical Image Analysis, 38, 104–116. Elsevier B.V. Retrieved from http://dx.doi.org/10.1016/j.media.2017.03.002
23. Wang, D., Khosla, A., Gargeya, R., Irshad, H., & Beck, A. H. (2016). Deep Learning for Identifying Metastatic Breast Cancer, 1–6. Retrieved from http://arxiv.org/abs/1606.05718
24. Wang, X., Chen, H., Gan, C., Lin, H., & Dou, Q. (2018). Weakly Supervised Learning for Whole Slide Lung Cancer Image Classification. Pdfs.Semanticscholar.Org, (Midl), 1–10. Retrieved from https://pdfs.semanticscholar.org/35d0/998f2c5b53591073d36c9e2b0ddc89a496b1.pdf
25. Wei, J. W., Tafe, L. J., Linnik, Y. A., Vaickus, L. J., Tomita, N., & Hassanpour, S. (2019). Pathologist-level classification of histologic patterns on resected lung adenocarcinoma slides with deep neural networks. Scientific reports, 9(1), 3358.
26. Xu, G., Song, Z., Sun, Z., Ku, C., Yang, Z., Liu, C., Wang, S., et al. (2019). CAMEL: A Weakly Supervised Learning Framework for Histopathology Image Segmentation. 2019 IEEE/CVF International Conference on Computer Vision (ICCV) (pp. 10681–10690). IEEE. Retrieved from https://ieeexplore.ieee.org/document/9008367/