The Histological Diagnosis of Colonic Adenocarcinoma by Applying Partial Self Supervised Learning

Abstract

Background: The cancer of colon is one of the important cause of morbidity and mortality in adults. For the management of colonic carcinoma, the definitive diagnosis depends on the histological examination of biopsy specimens. With the development of whole slide imaging, the convolutional neural networks are being applied to diagnose colonic carcinoma by digital image analysis. Aim: The main aim of the current study is to assess the application of deep learning for the histopathological diagnosis of colonic adenocarcinoma by analysing the digitized pathology images. Materials & Methods: The images of colonic adenocarcinoma and non neoplastic colonic tissue have been acquired from the two datasets. The first dataset contains ten thousand images which were used to train and validate the convolutional neural network (CNN) architecture. From the second dataset (Colorectal Adenocarcinoma Gland (CRAG) Dataset) 40% of the images were used as a train set while 60% of the images were used as test dataset. Two histopathologists also evaluated these images. In this study, three variants of CNN (ResNet-18, ResNet-34 and ResNet-50 ) have been employed to evaluate the images. Results: In the present study, three CNN architectures(ResNet-18, ResNet-30, and ResNet-50) were applied for the classification of digitized images of colonic tissue. The accuracy (93.91%) of ResNet-50 was the highest which is followed by ResNet-30 and ResNet-18 with the accuracy of 93.04% each. Conclusion: Based on the findings of the present study and analysis of previously reported series, the development of computer aided technology to evaluate the surgical specimens for the diagnosis of malignant tumors could provide a significant assistance to pathologists.

A Machine Learning Approach to Diagnosing Lung and Colon Cancer Using a Deep Learning-Based Classification Framework

“…Among the other studies, the authors of [ 28 ] worked with a set of colonoscopy images, and [ 31 , 34 ] worked based on CT scan images, so straightforward comparisons cannot be made. Only the studies cited in [ 51 , 52 , 53 ] work with the LC25000 dataset. Among them, reference [ 51 ] involves only the colon samples of the dataset and reported lower accuracy and F-measure scores than the proposed method.…”

“…Only the studies cited in [ 51 , 52 , 53 ] work with the LC25000 dataset. Among them, reference [ 51 ] involves only the colon samples of the dataset and reported lower accuracy and F-measure scores than the proposed method. Although [ 52 , 53 ] reported higher accuracy scores, they performed classifications either on the lung samples (three-class classifications) or on the colon samples (binary classifications).…”

Malignancy Detection in Lung and Colon Histopathology Images Using Transfer Learning With Class Selective Image Processing

“…Classification accuracy achieved by the model was 96.2%. S. Bukhari et.al, [25] applied three pre-trained CNNs (ResNet-50, ResNet-34, and ResNet-18) to evaluate the histopathology images of colonic adenocarcinoma from two different databases and acquired 96.4% accuracy. S. Mangal, A. Chaurasia, and Khajanchi [26] analyzed digital pathology images of adenocarcinoma and squamous cell carcinoma of the colon and lung and trained a shallow neural network for its classification, and achieved an accuracy of 97.8%.…”

“… Multiple datasets with varying scan settings, resulting in false-positive results [17].  Worked on a smaller number of classes or a less diverse dataset [20,21,25,26,27].  Reported low accuracy [11,12,14,15,20,22] The major contributions of this paper are outlined below: • Most earlier cancer detection studies focused on a single form of cancer, however in this study, we used our model to identify lung and colon cancer at the same time.…”

Multi-Input Dual-Stream Capsule Network for Improved Lung and Colon Cancer Classification

“…There have been a few important contributions to the classification of colon cancer. To classify histopathological images of colonic tissue, Bukhari et al [48] use three architectures of convolutional neural networks: ResNet-18, ResNet-30, and ResNet50. They claim that the models ResNet-30 and ResNet-18 each achieve 93.04% accuracy, while ResNet-50 achieves 93.91% accuracy.…”