Quantitation of glutamate in the brain by using MR proton spectroscopy at 1.5 T and 3 T

Purpose: The influence of different magnetic field strengths on the quantification of glutamate was experimentally investigated by means of in vitro and in vivo H-1-MR spectroscopic measurements at 1.5 T and 3 T. Materials and Methods: In vitro 1H-MR measurements of aqueous solutions of NAA, glutamate, glutamine and GABA were performed on two clinical MR scanners at 1.5 T and 3 T using a single voxel PRESS sequence (TR/TE=10000/ 30 ms). In vitro brain measurements were also performed at both field strengths using a PRESS 2D-H-1-CSI-sequence (TR/TE=5000/30ms) in 6 volunteers. Spectra at 1.5T and 3T were compared with respect to the overlap of the single compound spectra and the deviations between estimated and nominally adjusted concentrations. In vivo spectra at both field strengths were compared with respect to SNRGlu, line width and Cramer-Rao values of the estimated glutamate intensities by using the LCModel. For the thalamus, insular and parietal cortex mean Glu/tCr ratios were estimated and compared between 1.5T and 3 T as well as with corresponding values in the literature. Results: In general, an improved separation of signal maxima was observed in the in vitro spectra at 3 T. Except for GABA, all in vitro concentrations estimated at 3 T revealed lower deviations from their adjusted nominal concentration compared to 1.5 T: NAA (1.5 T: -5.5 %, 3 T: 0.7 %), glutamate (1.5T: -18.1%, 3T: 12.3%), glutamine (1.5T: 44.8%, 3T: 9.2%), GABA (1.5T: -24.8%, 3 T: 33.8 %). The SNR of in vivo spectra at 3 T was nearly doubled compared to 1.51 The mean number of voxels with %SDGlu< 20 was distinctly lower at 1.5 T (53 %) than at 3 T (80%). Estimated Glu/tCr ratios for thalamus, insular and parietal cortex lay in the upper range of the literature values. Conclusion: The results indicate that the advantageous distribution of signal maxima at 3T allows an improved separation of the individual spectra. Both the higher initial magnetization at 3 T and the improved sensitivity of the phased array matrix coil used in the 3T study result in an increased SNR, which leads to better reliability of the individual detection as well as a more accurate quantification of glutamate.