Optimization of the Use of Differences in FiO2 for the Study of Blood Perfusion in Magnetic Resonance Imaging

Brain tumors and ischemic diseases are one of the most aggressive causes of death. The study of cerebral perfusion, which is the passage of arterial blood from the circulatory system to the tissues, plays a key role in the diagnosis of these threatening medical conditions. Currently, cerebral perfusion is commonly studied using magnetic resonance imaging (MRI), which is a non-ionizing image modality. Perfusion in MRI is assessed in different ways; Dynamic Susceptibility Contrast (DSC) is the more wide spread method. DSC consists on injecting a bolus of an external contrast in blood, which in most cases is gadolinium, a paramagnetic contrast agent that facilitates transversal and longitudinal relaxation of the surrounding tissue. Therefore, the MRI signal varies when the contrast goes through the vessels. These changes in signal can be seen in T2/T2* weighted images, where the pass of gadolinium decreases the received signal. The shape of the intensity curves behaves differently in pathogenic tissue.

The main disadvantage of this procedure is precisely the use of this external contrast. It involves higher cost, discomfort and health risks for patients. Furthermore, this method does not rely on the intrinsic properties of blood inducing error in the measures. This project tries to solve this problem by using blood as an endogenous contrast for DSC. As Blood Oxygenated Level Dependent (BOLD) effect describes, deoxyhemoglobin is paramagnetic as gadolinium, so its higher concentration facilitates the transversal and longitudinal relaxation of spins. Therefore, decreasing the Fraction of Inspired Oxygen (FiO2) of a subject for a brief period of time, may lead to similar results to those obtained with gadolinium experiments. This project studies the relationship between changes in FiO2 and MRI signal. The MR images can be obtained using Echo Planar imaging (EPI) sequences which is a rapid imaging technique used in DSC. In this work the results obtained performing dynamic susceptibility contrast using changes in fraction of inspired oxygen, by forcing 100% nitrogen inspired gas (N2-DSC), were compared with gadolinium experiments (Gd-DSC) in both healthy and stroke-induced rats. Besides, a  MATLAB code was developed to generate the parametric maps of the perfusion parameters and to assess if comparable results between N2-DSC and Gd-DSC were obtained.