Retrospective analysis of quantitative imaging biomarkers as predictors of brain tumor clinical outcome and pathology

Although advanced imaging techniques provide quantitative, metabolic and functional data about brain tumors, it is evident that, as yet, there is no single technique that can offer a complete picture of brain tissue and tumor characterization. By generating a database of all brain tumor patients and their radiological and pathological studies and genetic biomarkers, we perform hypothesis-driven studies to assess whether biomarkers correlate with brain tumor pathology, grading, genetic phenotype, treatment response and patient outcome. The machine-learning and texture analysis techniques of the imaging are being used for grading and identifying molecular alteration and genetic profiles (radiogenomics) of gliomas.