Date of Award

Spring 4-20-2026

Document Type

Thesis

Degree Name

Master of Science (MS)

Department

Biomedical Engineering

First Advisor

Adam Greenberg

Second Advisor

Laura Glass Umfleet

Third Advisor

Robert Cooper

Abstract

As cancer survivorship increases, so does the prevalence of reported cognitive changes among patients. These changes have been attributed to multiple factors, including cancer treatments such as chemotherapy, radiation, and surgery, as well as the psychological stress associated with diagnosis. Prior research has identified potential predictors of cognitive change, including age, cognitive reserve, treatment regimen, and genetic factors, using approaches such as subjective assessments, neuropsychological testing, and neuroimaging. This study integrates large-scale behavioral data from the Health and Retirement Study (HRS; n = 310) with longitudinal task-based fMRI data from a breast cancer cohort undergoing chemotherapy (n = 9). Statistical modeling and feature selection techniques were applied to the HRS dataset to identify predictors of cognitive change. fMRI data were collected using the Attention Network Test (ANT) to define ROIs, and graph theory metrics were used to quantify functional connectivity for the task activated networks. Feature selection identified cancer treatment variables (chemotherapy, surgery, radiation), baseline cognition, and years of education as key predictors of cognitive outcomes. Neuroimaging results revealed connectivity changes in the alerting and orienting attention networks, measured using small-worldness. Graph theory analysis demonstrated a significant increase in betweenness centrality in the right superior parietal lobule (mSPL) within the orienting network (p = 0.02) following chemotherapy. Additionally, peripheral node centrality significantly increased in the alerting network (p= 0.04). These findings provide converging behavioral and neuroimaging evidence that cancer treatment is associated with cognitive changes and corresponding alterations in brain network connectivity.

Available for download on Thursday, May 04, 2028

Included in

Neurosciences Commons

COinS