A major problem in quantification of proton metabolites obtained by high field magnetic resonance (MR) spectroscopic imaging (MRSI) at short TE is separating the metabolite resonances from the uneven and varying baseline Some of the baseline fluctuations are due to out-of-volume water and lipid signals, while other variations are due to the presence of macromolecules We propose a series of measures to improve the ability to separate metabolite and baseline signals, the most important of which involves obtaining a series of moderate signal-to noise (S/N) spectra at differing TE times with a sequence that includes closely spaced 180 degrees pulses (CP pulse train) to reduce J evolution for strongly coupled spins prior to acquisition This preserves the spectral patterns for strongly coupled resonances at the different TE values even though the signal amplitudes decay due to T2 relaxation Simultaneous fitting of the suite of spectra is expected to enable improved separation of metabolite signals from baseline, and to provide more reliable estimates of metabolite resonance areas A number of additional measures including 1) improved RF pulses, including improvement of hyperbolic secant pulses and a spin echo pulse cascade with immunity to B1 inhomogenicity for the CP pulse train, 2) improved methods of shimming, 3) improved MRSI acquisition sequences, 4) improved methods for processing of metabolite spectra, and 5) improvements for metabolite quantification in institutional units, based on coanalysis with tissue segmented structural MRI data, are planned Finally, a metabolite atlas of normal brain, reflecting regional metabolite levels and variations, will be developed This proposal is coordinated with the partner IRPG proposal for sharing of programs, pulses, data, and information to enhance the research at both sites.
Showing the most recent 10 out of 11 publications