Item

Abstract

Proton magnetic resonance spectroscopy (1H MRS) in the human brain is a non-invasive
technique capable of aiding the investigation of the neurochemical composition. The
clinical importance of 1H MRS can be seen in pathological diagnosis, understanding
disease mechanisms or in treatment monitoring.
Reliable detection and quantification of metabolites is of paramount importance in
establishing potential biomarkers for several neurological pathologies. Furthermore,
broad macromolecular resonances underlying metabolite peaks in a proton spectrum also
hold a wealth of information. These macromolecular resonances originate from amino
acids within cytosolic peptides and proteins. Some studies in the past have even
discussed their clinical relevance in pathologies such as acute multiple sclerosis, glioma,
and traumatic encephalopathy. However, the characteristics of these macromolecular
resonances are yet to be fully explored. In-depth knowledge about the macromolecules
could open up a new horizon of potential biomarkers for neurological diseases. In
addition, characterizing macromolecular resonances may help the MR community answer
some of the lingering research questions such as identifying the biological background of
the individual macromolecular peaks, assigning macromolecular peaks to particular
amino acids, and investigating other contributions to the macromolecular signal such as
sugars, DNA or RNA.
Detection capabilities of MRS have increased to a great extent with increasing static
magnetic field. Ultra-high field (≥7 T) MRS benefits from increased signal-to-noise ratio
(SNR) and improved spectral resolution. There is also constant development in
localization techniques and quantification methods to accurately measure concentrations
of metabolites and macromolecules with lower signal-to-noise ratio and complex spectral
pattern due to J-coupling.
The first part of the thesis focuses on characterizing the physical properties of
macromolecular resonances in the human brain at 9.4 T and understanding their
contribution to the metabolite spectrum. T2 relaxation times are calculated and a
quantitative linewidth analysis is performed to understand the degree of overlap and J-coupling effects in the observed macromolecular peaks. Moreover, a novel double
inversion recovery method is proposed to determine T1 relaxation times of individual
macromolecular resonance lines.
The second part of the thesis focuses on quantification of metabolites in the human brain
at 9.4 T using one-dimensional and two-dimensional MRS techniques. Metabolite
concentrations are reported in millimoles/kg after correcting for T1- and T2-weighting
effects and the tissue composition. The concentration values measured from both the acquisition techniques were compared against each other and to literature.