Acquisition and Analysis of MRS Spectra in Animal Models

Proton Magnetic Resonance Spectroscopy (proton MRS) is a technique which allows to obtain biochemical information of several pathologies by studying the metabolite concentrations of the analyzed tissue. The strongest signal contribution is derived from the water, and it has to be removed during the online acquisition and the off-line post processing step. The presence and the removal of this huge signal may affect significantly the metabolites peaks and the baseline. Moreover acquisition of MRS spectra in living subjects can be affected by motion artifacts, bad shimming and low signal to noise ratio. Different post -processing techniques have been proposed to improve spectra obtained at difficult acquisition conditions, for water signal suppression and for baseline artifact correction. Nevertheless, among all the techniques proposed, it is difficult to identify the best performance and to rely on it, thus MRS quantification is still a challenge in biomedical research and not yet considered as a stable and fast routine diagnostic technique. In the present work, a comparison among the most used techniques in MRS quantification is proposed in two steps. In the first part of the thesis, different analysis technique have been applied on simple MRS signals, such as lipid spectra obtained from simulation and acquired in vivo and in vitro measurements. These spectra are characterized by a flat baseline and 11 to 12 slightly overlapping peaks, which are acquired without water suppression, so they constitute a simple field for comparison of quantification algorithms. In the second part of the thesis, these different techniques will be applied in the quantification of MRS of brain tissue. These spectra are much more complicated than lipid spectra in term of baseline and number of signals to discriminate. Under these experimental conditions, we will also compare methodologies in preclinical models of brain pathologies. Finally, the aim is to establish the best experimental procedures for both acquisition and analysis of in vivo brain spectra, to be applied first in preclinical, and subsequently in clinical research. An overview of the biological findings in both lipid and brain experimental projects will also be presented.