Analysis of Radiation Therapy in Cancer Treatment using Machine Learning

Yarschenko, Adam H

Analysis of Radiation Therapy in Cancer Treatment using Machine Learning

dc.contributor.advisor	Sun, Qiao
dc.contributor.advisor	Smith, Wendy L.
dc.contributor.author	Yarschenko, Adam H
dc.contributor.committeemember	Kirkby, Charles
dc.contributor.committeemember	Xue, Deyi
dc.contributor.committeemember	Smith, Wendy
dc.contributor.committeemember	Sun, Qiao
dc.date	Fall Convocation
dc.date.accessioned	2022-11-15T17:43:36Z
dc.date.embargolift	2022-09-24
dc.date.issued	2021-09-24
dc.description.abstract	Advancements in machine learning and data science have allowed researchers and clinicians to generate key insights from the vast amount of data generated in healthcare. This is currently a topic of research with great interest. With the advancement in algorithm design, and computing power, machine learning has proven to be a capable tool to augment or partially automate decision making. In this thesis, patient reported outcome surveys (PROs) for head and neck radiotherapy, and the relationship between the radiation dose distribution and breast size for whole-breast radiotherapy were investigated using statistical and machine learning methods. Two PRO measures; the M.D. Anderson symptom inventory for head and neck cancer, and the M.D. Anderson dysphagia inventory, were examined for a cohort of patients post radiotherapy for head and neck cancer. A strategy for administering a single PRO instrument is proposed which would reduce the questionnaire burden on patients, and allow physicians to identify patients who require specialized treatment for dealing with radiotherapy side-effects, such as referral to a dietician or speech language pathologist. Dosiomic features were extracted from the 3D radiation dose cloud in whole-breast radiotherapy plans. Feature reduction was achieved through hierarchical clustering and random forests were trained to stratify treated volume based on the distribution in the dose. Permutation feature importance was used to rank features’ classification utility in this task. The top 3 features were used to achieve superior performance when compared to the entire feature set. Dosiomics gives new insight into 3D dose distribution, and these features can be used in future studies to relate to treatment outcomes associated with whole breast radiotherapy for large volumes.
dc.identifier.citation	Yarschenko, A. H. (2021). Analysis of Radiation Therapy in Cancer Treatment using Machine Learning (Master's thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca.
dc.identifier.uri	http://hdl.handle.net/1880/115499
dc.identifier.uri	https://dx.doi.org/10.11575/PRISM/40466
dc.language.iso	en	en
dc.language.iso	English
dc.publisher.faculty	Graduate Studies	en
dc.publisher.faculty	Schulich School of Engineering
dc.publisher.institution	University of Calgary	en
dc.rights	University of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission.	en
dc.subject	radiation therapy
dc.subject	MADI
dc.subject	MDASI
dc.subject	machine learning
dc.subject	clustering
dc.subject	random forest
dc.subject	cancer
dc.subject	radiotherapy
dc.subject	patient reported outcomes
dc.subject	dosiomics
dc.subject.classification	Artificial Intelligence
dc.subject.classification	Engineering--Biomedical
dc.subject.classification	Physics--Radiation
dc.title	Analysis of Radiation Therapy in Cancer Treatment using Machine Learning
dc.type	master thesis
thesis.degree.discipline	Engineering – Mechanical & Manufacturing
thesis.degree.grantor	University of Calgary	en
thesis.degree.grantor	University of Calgary
thesis.degree.name	Master of Science (MSc)