High-grade gliomas are relatively rare and highly aggressive tumors affecting the brain. Despite significant research efforts in the last decades, survival for patients with high-grade gliomas is still poor. A number of research outputs have underlined the role of the hypoxic microenvironment on the aggressiveness and resistance to conventional therapies. Several hypoxia-modifying treatments are now available. However, there is an unmet need to bring a non-invasive imaging technique to the clinical setting in order to quantify tumor hypoxia and stratify patients with brain tumors based on hypoxia imaging findings. Oxygen-enhanced Magnetic Resonance Imaging (OE-MRI) appears as an intriguing and promising method to measure tissue hypoxia. The main goal of the two first experimental chapters of this thesis was to investigate the measurability and reproducibility of the percentage change of the longitudinal relaxation rate (R1), a commonly used parameter in OE-MRI studies to quantify tissue oxygenation, in normal brain and brain tumors. In a further study, employing Dynamic contrast-enhanced MRI, we explored the possibility of characterizing different tumor habitat types in patients with high-grade glioma and predict overall survival (OS) by means of the analysis of the histogram of the extravascular extracellular volume fraction (ve).