CEREBRAL ANATOMICAL AND FUNCTIONAL STUDIES THROUGH NUCLEAR MAGNETIC RESONANCE

This thesis introduces the results of the research done during the doctorate activities, and consists in an in-depth study on important issues of anatomical and functional brain imaging techniques through Nuclear Magnetic Resonance. The main objectives of the research can be summarized as a theoretical characterization of the MR signal produced by venous blood, an analysis and improvement of the techniques for high-resolution brain venography, and an investigation about the venous origins of the functional BOLD signal. After an introduction describing the basic principles of Magnetic Resonance Imaging and some important applications in the clinical routine, the first part of this work introduces a study in which the results of theoretical simulations on the MR venous signal are used to infer about the efficacy of a particular angiographic imaging methodology, such as high-resolution BOLD venography with T1-reducing contrast agent. An analytical model is described, through which the venous signal can be estimated starting from the vessel configuration. This can be described by several parameters that identify a generic vessel, like vessel size and inclination, position in the voxels, blood oxygenation, main magnetic field intensity, contrast agent concentration and so on. In the second part, a new methodology to perform BOLD venography is described, which represents a significant improvement of the technique. Starting from a detailed analysis on the performances of venographic imaging, pitfalls and further potentialities of the method are evidenced through theoretical and experimental considerations. As a result, a new data processing procedure is introduced which, at the same time, exploits at its best BOLD contrast for venographic imaging and provides a more efficient and flexible clinical tool. The validity of the method is demonstrated by tests done on digital phantoms and MR brain images. Finally, the last part of the thesis is dedicated to the investigation of the functional signal, with special regards to its origin and relation with brain vasculature. Functional experiments were performed on healthy subject contemporarily to high resolution anatomical scan, in order to correlate the functional response intensity with the underlying vascular structures. It was demonstrated experimentally that many functionally activated areas are found in correspondence with brain micro- and macrovasculature, and that the response intensity is directly related the vessel size. These results represent an experimental characterization of the BOLD functional signal, as well as the basis for possible filtering of venous artefacts in brain functional maps.