Classification of Epithelium Regions for Cervical Intraepithelial Neoplasia using Deep Learning
Department
Electrical and Computer Engineering
Major
Electrical Engineering and Mechanical Engineering
Research Advisor
Stanley, R. Joe
Advisor's Department
Electrical and Computer Engineering
Funding Source
Department of Electrical and Computer Engineering
Abstract
The scope of this OURE Fellows project is to investigate deep learning and big data techniques to detect key features in the epithelium in digitized histology images and to classify epithelium regions for Cervical Intraepithelial Neoplasia (CIN) discrimination. William Ong will be asked to label key features in the epithelium region from a database of over 200 digitized histology images to be used as inputs for deep learning algorithms. William will use existing image annotation tools available in Dr. R. Joe Stanley’s laboratory as well as baseline deep learning methods developed in Python for feature and CIN discrimination analysis. He will work with Dr. Stanley and also with Dr. William Van Stoecker for the image labeling process. William will apply and utilize these labeled features, including nuclei, cellular network and data fusion techniques to identify and characterize key features and to classify the epithelium region. All work will be done under the supervision of Dr. R. Joe Stanley, the faculty advisor for this project and will be mentored by two of Dr. Stanley’s Ph.D. graduate students.
Biography
William Ong is currently a student at Missouri S&T studying electrical and computer engineering, emphasizing in deep learning and AI. This past year, William had participated in Dr. Donnell’s Applied Microwave Thermography Nondestructive Testing lab. William also has participated in Formula SAE and the IEEE student branch. William’s interests include biology and big data analytics. William is passionate in using AI and deep learning to aid people in meaningful ways.
Presentation Type
OURE Fellows Proposal Oral Applicant
Document Type
Presentation
Award
2016-2017 OURE Fellows recipient
Location
Turner Room
Presentation Date
17 Apr 2018, 2:00 pm - 2:20 pm
Classification of Epithelium Regions for Cervical Intraepithelial Neoplasia using Deep Learning
Turner Room
The scope of this OURE Fellows project is to investigate deep learning and big data techniques to detect key features in the epithelium in digitized histology images and to classify epithelium regions for Cervical Intraepithelial Neoplasia (CIN) discrimination. William Ong will be asked to label key features in the epithelium region from a database of over 200 digitized histology images to be used as inputs for deep learning algorithms. William will use existing image annotation tools available in Dr. R. Joe Stanley’s laboratory as well as baseline deep learning methods developed in Python for feature and CIN discrimination analysis. He will work with Dr. Stanley and also with Dr. William Van Stoecker for the image labeling process. William will apply and utilize these labeled features, including nuclei, cellular network and data fusion techniques to identify and characterize key features and to classify the epithelium region. All work will be done under the supervision of Dr. R. Joe Stanley, the faculty advisor for this project and will be mentored by two of Dr. Stanley’s Ph.D. graduate students.
