Advanced metabolite mapping at ultra-high field using 1H-MRS, 1H-MRSI and macromolecules: applications in a rat model of type C hepatic encephalopathy

Magnetic Resonance Spectroscopy (MRS) is the only technique capable of measuring a large number of metabolites simultaneously in vivo. Ultra-high magnetic fields (UHF) combined with ultra-short echo time (TE) sequences allow the detection of high-quality 1H MR spectra and the quantification of 20 different metabolites in the brain (markers of energy metabolism, osmoregulation etc.).In vivo brain localized 1H MR spectra at short TEs contain the contribution of mobile macromolecules (MM). Reliable detection and fitting of MM are crucial for accurate quantification. Higher spectral resolution at UHF led to increased interest in using a parametrized MM spectrum and flexible spline baselines to address unpredicted spectroscopic components.In this thesis the MM spectra (from the rat brain at 9.4T) were characterized using an improved methodological approach for their post-processing, fitting and quantification. This method provided an efficient tool for parametrization of the MM spectrum into individual components and estimation of their T2app relaxation times. An extensive assessment on how the MM spectrum and spline baseline stiffness affect the metabolite and MM quantification is also reported Type C hepatic encephalopathy (HE) is a complication of chronic liver disease (CLD). Children and adults respond differently to CLD and its related toxic accumulation of molecules (i.e. ammonium (NH4+), glutamine (Gln)). Children with CLD may grow up with significant neurocognitive deficits even after liver transplantation. Despite considerable advances in understanding the pathogenesis of type C HE, the exact metabolic mechanisms and their regional variations are not fully understood.The advantages of UHF short TE 1H MRS were used herein to describe the regional distribution of metabolites in the developing and adult brain using the bile duct ligated model (BDL) of type C HE (adult and postnatal day 21 rats). Three brain regions were assessed (hippocampus, cerebellum and striatum) pointing towards cerebellum as a region with the heaviest burden of Gln and unique metabolic response. Changes in cell morphology were followed longitudinally and related to the metabolic alterations. Elevated oxidative stress is reported using electron paramagnetic resonance, together with the decreased antioxidants (1H MRS) emphasizing its important role in HE. The brain regional measurements confirmed the higher susceptibility of developing brain to the disease and the increased vulnerability of cerebellum. Finally, the beneficial effect of Cr supplementation on the neurometabolic profile is described using 1H MRS and 31P MRS in CLD pups (BDL at postnatal day 15) suggesting that an appropriate treatment may have significant public health impact.MRSI is a powerful tool to non-invasively and spatially map the brain regional distribution of metabolites in vivo. While MRSI in human brain is increasingly used, preclinical MRSI is not widely applied mainly due to the small rodent brain, long acquisition times and low signal to noise ratio. The implementation of a novel approach: free induction decay (FID) MRSI on the 14.1T preclinical scanner is described herein. This method offers a fast and robust data acquisition with high spatial resolution resulting in high quality spectroscopic maps. Finally, preliminary assessment of the effect of two noise reduction techniques (MP-PCA and TGV reconstruction) on the spectra from preclinical MRSI datasets is briefly presented.