In Vivo Assessment of Human Cerebral Glucose Oxidative Metabolism Using 1H and 13C MRS at 7T: Toward Clinical translation

Abnormal cerebral energy metabolism is closely associated with many neuropsychiatric and neurodegenerative disorders. Carbon-13 magnetic resonance spectroscopy (MRS), combined with labeled substrate infusion, offers a unique means to measure brain metabolic fluxes; however, its clinical application has been limited by low intrinsic sensitivity, the need for blood sampling, specialized hardware requirements, high radiofrequency power deposition at ultra-high field, and the complexity of spectral modeling. This thesis presents a methodological framework for developing and validating a clinically translatable 7-Tesla interleaved 1H/13C MRS protocol for minimally invasive, in vivo quantification of cerebral glucose uptake, downstream metabolic pathways, and tricarboxylic acid (TCA) cycle flux in the frontal lobe. Interleaved 1H/13C MRS sequences, including ACE-STEAM and ISIS-DEPT, were optimized for reliable detection of key metabolites such as 13C-labeled glucose, glutamate, and glutamine. A custom 3/2-channel 1H/13C RF surface coil was designed to maximize frontal lobe coverage and signal sensitivity. Furthermore, a spectral processing software platform, MRSpecLAB, was developed to simplify the spectral analysis pipeline and perform automated data processing. Linewidth-matched (LM) basis sets in LCModel were investigated to enhance quantification accuracy. The integrated protocols were validated for estimating metabolic fluxes, demonstrating feasibility for probing dynamic brain glucose metabolism and establishing a foundation for translational clinical applications at 7 Tesla.