Cervical Histopathology Image Classification Using Multilayer Hidden Conditional Random Fields and Weakly Supervised Learning

Abstract

In this paper, a novel multilayer hidden conditional random fields (MHCRFs)-based cervical histopathology image classification (CHIC) model is proposed to classify well, moderate and poorly differentiation stages of cervical cancer using a weakly supervised learning strategy. First, the color, texture, and deep learning features are extracted to represent the histopathological image patches. Then, based on the extracted features, artificial neural network, support vector machine, and random forest classifiers are designed to calculate the patch-level classification probabilities. Third, effective classifiers are selected to generate unary and binary potentials. At last, using the generated potentials, the final image-level classification results are predicted by our MHCRF model, and an overall accuracy around 77.32% is obtained on six practical cervical histopathological image datasets with more than 600 immunohistochemical (IHC) stained samples. Among the six test accuracies, the highest reaches 88%. Furthermore, we also test our MHCRF method with a gastric hematoxylin-eosin (HE) stained histopathological image dataset including 200 images for an extended experiment, and achieve an accuracy of 93%.

An Application of Transfer Learning and Ensemble Learning Techniques for Cervical Histopathology Image Classification

“…The accuracy of the ETL method [13] is 2.7% higher than the single TL method. However, our proposed method achieves the highest accuracy of 98.61% among all the methods, and the accuracy is 10.61% higher than the second [11], [12], showing the effectiveness of our ETL method in this paper.…”

“…In another work [11], we suggest a multilayer hidden conditional random fields (MHCRFs) to classify well, moderate and poorly differentiation stages of cervical cancer, and an accuracy of 88% is obtained on a practical histopathological image dataset with more than 100 AQP stained samples. Meanwhile, in [12], a novel MHCRFs based cervical histopathology image classification model is proposed to classify well, moderate, and poorly differentiated stages of cervical cancer using a weakly supervised learning strategy. In [29], we utilize graph and unsupervised learning methods in a tissue structure clustering task, and divide the histopathological images of cervical cancer into sparse areas and viscose areas to predict the risk of the tissues.…”

A Comprehensive Review for Breast Histopathology Image Analysis Using Classical and Deep Neural Networks

“…The scope of the application focuses on the benign and malignant diagnosis, disease grading, staining analysis, and early tumor screening. For example, the work of [13] proposes a weakly supervised multi-layer hidden conditional random field model to classify the cervical histopathological images into well, moderate and poorly differentiated stages. In the experiment, the proposed method is tested on the six cervical IHC datasets and obtains an overall classification accuracy of 77.32% and the highest one of the six is 88%, showing the effectiveness and potential of the method.…”

A Cervical Histopathology Image Clustering Approach Using Graph Based Features

“…In our previous work [16,18,19,46], an ensembled transfer learning framework is introduced to classify well, moderately and poorly differentiated cervical histopathology images. As well, in our previous work [14,15], a novel multilayer hidden conditional random fields based weakly supervised cervical histopathological image classification framework is proposed to classify three differentiation stages of cervical cancer.…”