Convolutional Neural Networks for Classification of Colorectal Cancer on Whole Slide Images

Temena, Mehmet Arda
Isildak, Ulas
Acar, Ahmet
Diagnosis of cancer is typically performed by medical pathologist visually inspecting hematoxylin and eosin (H&E) stained whole slide images. Although the early detection of various cancer types is crucial for a successful treatment, it is a challenging work due to inter- and intra-personal variability and it may often result in a disagreement between pathologists. Visual inspection of medical specimen is a well- established method for cancer diagnosis, but digital transformation is now inevitable fact, whereby conventional microscopic examinations with bareeyes have been replaced by artificial learning (AI)- based digital pathology applications in last years. For this reason, AI-driven solutions are now helping to operate on biospecimen images mostly to assist pathologists. Moreover, deep learning-based diagnosis systems have showed a potential for reducing the cost and improving the accuracy. In this work, we aimed to develop a automated pipeline for colorectal cancer diagnoses by applying Convolutional Neural Network (CNN) based models to classify whole slide images in two classes, carcinoma and non-carcinoma. We included several previously developed deep learning classifiers in our pipeline, where a data set with more than 300 annotated and accessible whole-slide images for colorectal cancer will be used to evaluate model performances.


Automated cancer stem cell recognition in H&E stained tissue using convolutional neural networks and color deconvolution
Aichinger, Wolfgang; Krappe, Sebastian; ÇETİN, AHMET ENİS; Atalay, Rengül; ÜNER, AYŞEGÜL; Benz, Michaela; Wittenberg, Thomas; Stamminger, Marc; Muenzenmayer, Christian (2017-02-13)
The analysis and interpretation of histopathological samples and images is an important discipline in the diagnosis of various diseases, especially cancer. An important factor in prognosis and treatment with the aim of a precision medicine is the determination of so-called cancer stem cells (CSC) which are known for their resistance to chemotherapeutic treatment and involvement in tumor recurrence. Using immunohistochemistry with CSC markers like CD13, CD133 and others is one way to identify CSC. In our wor...
Mercadier, Deniz Sayin; Beşbınar, Beril; Frossard, Pascal (2019-01-01)
Accurate and fast segmentation of nuclei in histopathological images plays a crucial role in cancer research for detection and grading, as well as personal treatment. Despite the important efforts, current algorithms are still suboptimal in terms of speed, adaptivity and generalizability. Popular Deep Convolutional Neural Networks (DCNNs) have recently been utilized for nuclei segmentation, outperforming traditional approaches that exploit color and texture features in combination with shallow classifiers o...
Overexpressed genes/ESTs and characterization of distinct amplicons on 17q23 in breast cancer cells
Erson Bensan, Ayşe Elif; DEMERS, SK; ROUILLARD, JM; HANASH, SM; PETTY, EM (Elsevier BV, 2001-11-01)
17q23 is a frequent site of gene amplification in breast cancer. Several lines of evidence suggest the presence of multiple amplicons on 17q23. To characterize distinct amplicons on 17q23 and localize putative oncogenes, we screened genes and expressed sequence tags (ESTs) in existing physical and radiation hybrid maps for amplification and overexpression in breast cancer cell lines by semiquantitative duplex PCR, semiquantitative duplex RT-PCR, Southern blot, and Northern blot analyses. We identified two d...
Multiplication free neural network for cancer stem cell detection in H&E stained liver images
Badawi, Diaa; Akhan, Ece; Mallah, Ma'en; ÜNER, AYŞEGÜL; Atalay, Rengül; Cetin, A. Enis (2017-04-13)
Markers such as CD13 and CD133 have been used to identify Cancer Stem Cells (CSC) in various tissue images. It is highly likely that CSC nuclei appear as brown in CD13 stained liver tissue images. We observe that there is a high correlation between the ratio of brown to blue colored nuclei in CD13 images and the ratio between the dark blue to blue colored nuclei in H&E stained liver images. Therefore, we recommend that a pathologist observing many dark blue nuclei in an H&E stained tissue image may also ord...
Polyp Detection in Colonoscopy Images using Deep Learning and Bootstrap Aggregation
Polat, Görkem; Işık Polat, Ece; Kayabay, Kerem; Temizel, Alptekin (2021-4-13)
Computer-aided polyp detection is playing an increasingly more important role in the colonoscopy procedure. Although many methods have been proposed to tackle the polyp detection problem, their out-of-distribution test results, which is an important indicator of their clinical readiness, are not demonstrated. In this study, we propose an ensemble-based polyp detection pipeline for detecting polyps in colonoscopy images. We train various models from EfficientDet family on both the EndoCV2021 and the Kvasir-S...
Citation Formats
M. A. Temena, U. Isildak, and A. Acar, “Convolutional Neural Networks for Classification of Colorectal Cancer on Whole Slide Images,” presented at the 13th CONGRESS of MEDICAL INFORMATICS, Ankara, Türkiye, 2021, Accessed: 00, 2021. [Online]. Available: