Growing interest in the research community has been shown in clinical neuroscience to assess neurotransmitter profiling both in healthy and diseased subjects. A large body of research in this field focuses on schizophrenia to characterise its glutamatergic level according to the most recent hypothesis of NMDA (N-Methyl-D-aspartic acid) receptors hypofunction. Magnetic Resonance Spectroscopy (MRS) is able to detect some of the most common neurotransmitters but a number of issues, such as low signal to noise ratio (SNR), spectra overlapping and line broadening prevents MRS from being clinically relevant for neuropsychiatry. Four important aims were considered relevant for this work. Firstly, we aimed to compare the reliability of conventional and timing-optimized sequences for the detection and measurement of most of the visible metabolites and, in particular, for glutamate (Glu), glutamine (GIn) and gamma-aminobutyric acid (GABA) to assess the best available sequence for a study in schizophrenia. Secondly, we also intended to investigate whether glutamatergic activity might predict the oscillatory activity and how this link might survive or not in schizophrenia. Thirdly, we wanted to study whether the well known animal model of schizophrenia, the rearing in isolation model, exacerbates the effect of ketamine and determines more profound changes on neurotransmitter profile in rats. Fourthly, a further goal focuses on the improved data acquisition and on the data processing to reliably resolve GABA and to be able to quantify a wider range of metabolites. To address those points five studies were performed. The first work (Chapter 3) describes a study of reproducibility on sequences which have been reported in the literature to be capable to detect Glu and GIn. The study was performed on 14 healthy subjects by scanning them twice and repositioning between the two scans. The absolute percentage difference was then computed to assess the accuracy per sequence and metabolite. A good compromise was found in PRESS sequence (TE=80 ms) which was exploited subsequently for the following study on schizophrenic patients (Chapter 4). Twenty-seven early stage schizophrenic patients and twenty-three aged-matched controls were recruited to undergo a protocol including, in two separate sessions, MRS and electroencephalography (EEG). Anterior Cingulate Cortex Glu was found to predict the induced theta activity in healthy controls but not in patients. Furthermore, the NAA values have also been found to be reduced in schizophrenia and linked to N100, an Event Related Potential (ERP) which is well known to be decreased in schizophrenia. Following on from the findings of the study on the early stage of schizophrenia, further investigations were undertaken to study the psychotic state occurring in the disease via a functional MRS, where 25mg/kg of ketamine (NMDA antagonist) injection was administered to two groups of rats. The two groups were group-housed and reared in isolation. This work was able to show increase of prefrontal GIn levels in both groups but showed a selective GABA decrease only in isolated rats. It would have been very interesting to be able to detect GABA changes in the study at 3T but the used protocol did not allow its accurate quantification. Simulations and reliability tests (Chapter 6)were then utilized to optimize a standard sequence to obtain an accurate and reliable GABA concentration. The optimized sequence reproduces the quantification with 12% of accuracy. The preliminary results of the last study (Chapter 7) give an evidence of the potential of combined use of Monte Carlo, Levenberg-Marquardt and NNLS methods embedded in a novel fitting approach for two-dimensional spectra. The three appendices at the end of this work illustrate the details of some of the algorithms and softwares used throughout the studies.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:546535 |
| Date | January 2011 |
| Creators | Napolitano, Antonio |
| Publisher | University of Nottingham |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://eprints.nottingham.ac.uk/28305/ |
Page generated in 0.0023 seconds