Biophysical Simulation of the Functional Magnetic Resonance Signal Formation in Realistic Neurovascular Networks

Abstract

The human brain is one of the most complex living systems. The scientific study of the brain’s anatomy and neurophysiology are fundamental to understand the basic principles of mental processes such as cognition and behavior. For this reason, in the endeavor to investigate the underlying neural mechanisms that drive these processes, the neuroscientific research has employed all the available technological resources and methodologies from different fields like anatomy, histology, electrophysiology, neurobiology, etc. Likewise, great advances have been provided by neuroimaging techniques such as PET and MRI, in order to comprehend the neural activity and the metabolic reactions that occur in the central nervous system. In particular, functional MRI provides an indirect measurement of the neural activity throughout local hemodynamic changes, thus related to the neurovascular coupling, as a response to a particular task-evoked stimulus. This MR signal behavior modulated by oxygenation level changes is better known as the BOLD signal. Although progress has been done in order to understand the BOLD signal change under well-defined nonrealistic vascular geometries, on the other hand, realistic neurovascular networks might give valuable information to resolve the influence on the BOLD signal evolution from a particular vascular tissue and specific hemodynamic responses.

In order to extend the analysis of the BOLD signal change obtained by randomly oriented cylinders and spheres, throughout this thesis, the geometrical features of a realistic neurovascular network as well as the biophysical effects related to the hemodynamic response and thermal motion were investigated by means of Monte Carlo simulations in pursuance to resolve the functional MR signal formation. In the Introduction of this thesis, I made a small recapitulation on the MR physics and spin dynamics; magnetic susceptibility and thermal motion as crucial modulators of the BOLD signal behavior. In addition, a summary of the problem and the aims of the project. Therefore, I described the importance of the use of the Monte Carlo method to calculate the MR signal under nonrealistic vascular models. I summarized the seminal analytical and numerical results that provide important insights to characterize the main parameters that influence the MR signal formation. Finally, I described the importance of the use of realistic neurovascular structures in order to disentangle the specific tissue contribution and the direct impact on the BOLD signal change.