Ultra-short acquisition delay 3D Spectroscopic Imaging

Proton magnetic resonance spectroscopic imaging (MRSI) enables the non-invasive imaging of local changes in brain neurochemistry that underlie many brain diseases. Our goal is to develop an improved MRSI method for the robust whole brain-mapping of a comprehensive number of chemical compounds. Commonly available MRSI methods share several methodical limitations, which include the following: (a) a low spatial resolution; (b) limited brain coverage; (c) vulnerability to artifacts; (d) substantial problems with localization accuracy; (e) quantification of only a limited number of neurochemical compounds. To make MRSI a powerful clinical/neuro-scientific imaging tool, these limitations have to be overcome. Robust whole-brain-MRSI must provide accurate mapping of a comprehensive neurochemical profile with higher spatial resolution. Such MR imaging methods will be valuable tools for clinical and neuro-scientific investigations of various pathologies such as MS, tumors, or neurodegenerative diseases. We hypothesize that: (a) methods that accelerate data sampling can be combined to allow for high-resolution whole-brain MRSI within clinically feasible scan times; (b) real-time artifact correction will allow robust MRSI acquisition with reproducible results for clinical/neuro- scientific application; and (c) imaging of an extended number of neurochemical compounds will facilitate the understanding of complex neurochemical alterations underlying major brain pathologies. So far only one research group has established an MRSI sequence with whole brain coverage, but only for lower field strength and using methods that are less sensitive to most chemical compounds. No implementation for high field strengths exists, because of technical limitations. Based on our preliminary sequence development, we will develop a whole brain MRSI technique that: (a) can be used also at ultra-high field; (b) will allow mapping of more chemical compounds; (c) will be far more sensitive than previous approaches; (d) will allow a higher spatial resolution; and (e) will be more robust than available techniques. The successful completion of the project will, therefore, permit the development of a new inherently multi-parametric neuroimaging technique that can be a powerful tool for clinical and neuro-scientific investigations.

In this project we have developed a whole-brain imaging technique that is able to map the concentrations of multiple biochemical substances within the human brain in clinically attractive measurement times. This has a high potential for an array of clinical applications, because local or widespread changes in brain biochemistry are ubiquitous and important in many major brain disorders. However, the acquired raw data are enormous and consequently the processing times to turn these raw data into concentrations maps are still very long, which limit currently the use of these techniques for routine use. Nevertheless, we have performed initial clinical evaluations in a cohort of patients with brain tumors and Multiple Sclerosis with promising results. In follow-up studies we will focus on a better integration of these new neurochemical imaging techniques on the magnetic resonance scanners and improved robustness to enable a user-friendly application in a clinical environment.