Mario Baez

Biomedical Engineer
PhD Student at Max Planck Institute for Biological Cybernetics

Sequences and Signals, NMR and MRI Signal Formation, fMRI & Monte-Carlo Simulations

Monte-Carlo Simulation of GE, SE and bSSFP MR signal formation in Neurovascular Tissue at 9.4 T

Modelling the functional nuclear magnetic resonance (NMR) signal opens the possibility to go beyond the actual magnetic resonance imaging (MRI) resolution and extract the fingerprint mesoscopic structure immerse in an imaged voxel.

High resolution fMRI is one of the most interestingly imaging acquisition methods because it allows more accurate spatial mapping of brain responses. This type of mapping relies on the blood oxygen level dependent (BOLD) contrast, and is closely related with the dynamics of the neurovascular network, responsible of the hemodynamics exchanges. Moreover, high resolution fMRI start to reveal the vascular and metabolic heterogeneity of the cortex, since the relevant contributor structures are of a similar size as the image resolution. To accurately predict high resolution BOLD responses, the characterisitc features of the model and the NMR signal needs to be modeled explicitly.

Biological tissues have a complicated chemical composition and geometrical architecture that influence the MR signal formation. In the presence of an external magnetic field, differences in magnetic susceptibilities between adjacent compartments generate local magnetic field perturbations, and hence a range of magnetic resonance frequency shifts. In addition, water molecules are not static but diffuse freely or restricted within this local inhomogenous field pattern, including an extra-dephasing incoherence to the spin frequency.

The use of bSSFP acquisition sequence in a fMRI experiment and the relation of the MR signal at microscopic scales is poorly understood. Aiming to understand the brain function and its relation with the MR signal, computer simulations supply valuable information to clarify this MR signal formation under realistic tissue properties scheme in both health and disease.

Neurovascular structure acquired with two-photon imaging from the rat brain cortex. Cylinders in different orientations respect to the main magnetic field mimicking homogeneous neurovascular structures. The frequency distribution shift created by the de-/oxy hemoglobin states can be calculated by analytical methods. The molecular motion within the frequency distribution exhibit by the model can be modelated with random walkers.