Files
Abstract
Radiomics is an emerging field that focuses on expanding the role of human image interpretation to incorporate computer vision, artificial intelligence and machine learning. The goal is to gather statistical relationships of tumors as derived from various imaging modalities, including computed tomography (CT), and provide a deeper understanding of the properties that define a tumor in contrast with normal tissue. This could explain diagnostic and prognostic attributes of disease and predict an individual's unique response to treatment. The overall aim of this work was to develop a workflow for radiomic analysis and identify variables during biopsy procedures that could provide insight into the composition of a patient's tumor burden and predict response to treatment. The study found that several variables, including kurtosis, gray level variance, volume, sedation time, and procedure complications could play a role in determining prognosis.