Analyzing Magnetic Resonance Spectroscopy Data

Abstract

Magnetic resonance (MR) spectroscopy (MRS) modalities provide non-invasive and non-ionization in vivo imaging tools for preclinical and clinical examinations. Brain examinations’ standard clinical MR protocols comprise several anatomical imaging techniques. This work describes the principal physics behind clinical MRS, spectral analysis paradigm, and MRS applications in clinical routines. The excellent contrast from the brain’s anatomy partly relies on water’s hydrogen nuclei relaxation time differences in tissues. Peak fitting and a linear combination of simulated metabolites are standard algorithms to estimate metabolite intensities from MR spectra. This thesis aims to implement two popular algorithms on in vivo clinical MR spectra and compare the quantification estimations of two methods.

Magnetic resonance (MR) spectroscopy (MRS) modalities provide non-invasive and non-ionization in vivo imaging tools for preclinical and clinical examinations. Brain examinations’ standard clinical MR protocols comprise several anatomical imaging techniques. This work describes the principal physics behind clinical MRS, spectral analysis paradigm, and MRS applications in clinical routines. The excellent contrast from the brain’s anatomy partly relies on water’s hydrogen nuclei relaxation time differences in tissues. Peak fitting and a linear combination of simulated metabolites are standard algorithms to estimate metabolite intensities from MR spectra. This thesis aims to implement two popular algorithms on in vivo clinical MR spectra and compare the quantification estimations of two methods.