Date of Award
Summer 2021
Publication Type
Thesis
Degree Name
M.A.Sc.
Department
Industrial and Manufacturing Systems Engineering
Keywords
Biomarkers. Breast cancer. Classification. Lymph nodes. Machine learning. Neural network
Supervisor
W. ElMaraghy
Supervisor
A. Alkhateeb
Rights
info:eu-repo/semantics/openAccess
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Abstract
After a patient’s breast cancer diagnosis, identifying breast cancer lymph node metastases is one of the most important and critical factor that is directly related to the patient’s survival. The traditional way to examine the existence of cancer cells in the breast lymph nodes is through a lymph node procedure, biopsy. The procedure process is time-consuming for the patient and the provider, costly, and lacks accuracy as not every lymph node is examined. The intent of this study is to develop an artificial neural network (ANNs) that would map genetic biomarkers to breast lymph node classes using ANNs. The neural network classes were adopted from the American joint committee on cancer materials pathologic lymph node statuses. Twenty gene biomarkers were selected based on a hybrid feature selection model, and the neural network parameters were configured using hyperparameter tuning techniques targeting the neural network capacity, activation function, weight initialization, and the neural network learning rate and momentum. The METABRIC breast cancer data set was used to train, validate, and test the neural network. The results show an accuracy of 96% for the training dataset and 85% for both the validation and test dataset. As far as the area under the curve (AUC), the neural network scored 100% for the training dataset, 95% for the validation, and finally, 90% for the test dataset. This study directly benefits and supports the cancer organizations transition from identifying cancer based on the organ in the patient’s body, where cancer first starts to develop as well as the shape of cancer under a microscope, to grouping cancer cells based on gene mutations. This change in cancer identification will assist the providers in improving the diagnosis, prognosis, and treatment of cancer patients.
Recommended Citation
Omar, Ziad, "Deep Learning Applications in Medical Bioinformatics" (2021). Electronic Theses and Dissertations. 8829.
https://scholar.uwindsor.ca/etd/8829