Files
Abstract
This dissertation quantitatively assesses the single-cell spatial landscape of the tumor microenvironment for three cancer types. Statistical and machine learning approaches were used to identify candidate biomarkers of various clinical parameters, which can be leveraged to improve treatment strategies for future cancer patients. Importantly, the computational methods used here are widely applicable regardless of tissue type, and they provide a framework for future quantitative assessments of biological datasets resulting from any multiplex tissue imaging assay.