Magnetic resonance imaging (MRI) is a noninvasive technique that allows imaging three dimensional volume through the body, providing superior soft tissue contrast compared to CT without ionizing radiation. MRI can be used to detect field inhomogeneity induced by blood iron molecules such as deoxyhemoglobin, methemoglobin, and hemosiderin, providing opportunities to identify and quantify these molecules and oxygen consumption, especially in brain. This dissertation reports: 1) a clinical study on intracranial hematoma evolution using quantitative susceptibility mapping (QSM); 2) a new magnetic susceptibility based signal model and MRI technique to quantitatively map cerebral metabolic rate of oxygen (CMRO2) and oxygen extraction fraction (OEF) using QSM; 3) a new algorithm to improve SNR of QSM-based CMRO2 using prior knowledge; 4) a new algorithm called minimal local variance (MLV) to improve clinical practicality of QSM-based CMRO2 by removing the requirement of blood flow challenge. The potential impacts, limitations and future directions of each studies are also discussed.