Summary
- Major: Biomedical & Electrical Engineering
- Faculty Mentor: Dr Giulia Debiasi, University of CA, Berkeley
- Research Topic: Development of a Quantitative Susceptibility Mapping Pipeline for Identification of Iron
Related Biomarkers in Glioma
Abstract
Brain tumors, including gliomas, exhibit heterogeneous tissue composition and infiltration into surrounding brain regions, making accurate characterization challenging with conventional magnetic resonance imaging (MRI). We hypothesize that quantitative susceptibility mapping (QSM) can identify regional differences in magnetic susceptibility within tumor tissue and surrounding edema, providing quantitative measures of iron-related tissue changes that may improve characterization of tumor microenvironments. To investigate this hypothesis, we developed a computational pipeline for quantitative analysis of QSM data acquired from 28 retrospectively enrolled brain tumor patients MRI datasets. The workflow integrates MATLAB and Python-based processing to reconstruct susceptibility maps, segment tumor and edema regions, and quantify regional susceptibility measurements. Analyses were performed across multiple glioma subtypes, including astrocytoma (Type 1), oligodendroglioma (Type 2), and glioblastoma (Type 3), to compare susceptibility patterns among tumor types and surrounding peritumoral edema. NumPy, SciPy, and Matplotlib were used for image processing, quantitative analysis, and visualization of susceptibility distributions across anatomically defined regions of interest. Preliminary analyses demonstrate successful implementation of the computational pipeline and reproducible regional susceptibility measurements across patient datasets. Initial findings reveal measurable differences in magnetic susceptibility between tumor regions and surrounding edema, supporting the feasibility of QSM for characterizing spatial variations in tissue composition across glioma subtypes. This work establishes a reproducible framework for regional QSM analysis of brain tumors and provides a foundation for future studies investigating susceptibility-based imaging biomarkers across astrocytoma, oligodendroglioma, glioblastoma, and other brain tumor types. Improved quantitative characterization of tumor and peritumoral regions may ultimately enhance disease assessment,
treatment planning, and longitudinal monitoring.