Keywords and Phrases
Cervical cancer; Convolutional neural networks; Deep learning; Digital pathology; Histopathology; Whole slide image
“Cervical cancer is the fourth most frequent cancer that affects women worldwide. Assessment of cervical intraepithelial neoplasia (CIN) through histopathology remains as the standard for absolute determination of cancer. The examination of tissue samples under a microscope requires considerable time and effort from expert pathologists. There is a need to design an automated tool to assist pathologists for digitized histology slide analysis. Pre-cervical cancer is generally determined by examining the CIN which is the growth of atypical cells from the basement membrane (bottom) to the top of the epithelium. It has four grades, including: Normal, CIN1, CIN2, and CIN3. In this research, different facets of an automated digitized histology epithelium assessment pipeline have been explored to mimic the pathologist diagnostic approach. The entire pipeline from slide to epithelium CIN grade has been designed and developed using deep learning models and imaging techniques to analyze the whole slide image (WSI). The process is as follows: 1) identification of epithelium by filtering the regions extracted from a low-resolution image with a binary classifier network; 2) epithelium segmentation; 3) deep regression for pixel-wise segmentation of epithelium by patch-based image analysis; 4) attention-based CIN classification with localized sequential feature modeling. Deep learning-based nuclei detection by superpixels was performed as an extension of our research. Results from this research indicate an improved performance of CIN assessment over state-of-the-art methods for nuclei segmentation, epithelium segmentation, and CIN classification, as well as the development of a prototype WSI-level tool”--Abstract, page iv.
Stanley, R. Joe
Stoecker, William V.
Kosbar, Kurt Louis
Zawodniok, Maciej Jan, 1975-
Samaranayake, V. A.
Electrical and Computer Engineering
Ph. D. in Computer Engineering
Missouri University of Science and Technology
Journal article titles appearing in thesis/dissertation
- Deep learning nuclei detection in digitized histology images by superpixels
- EpithNet: Deep regression for epithelium segmentation in cervical histology images
- DeepCIN: Attention-based cervical histology image classification with sequential feature modelling for pathologist-level accuracy
- Automated cervical digitized histology whole slide image analysis toolbox
xviii, 125 pages
© 2020 Sudhir Sornapudi, All rights reserved.
Dissertation - Open Access
Sornapudi, Sudhir, "Deep learning for digitized histology image analysis" (2020). Doctoral Dissertations. 3110.