hide
Free keywords:
-
Abstract:
Purpose/Introduction: In many MRS measurements metabolite
intensities are referenced to assumed amounts of internal water (IWR)1 or total Creatine (tCr). However, since the water and tCr concentrations as well as the relaxation properties differ between grey matter (GM), white matter (WM) and cerebrospinal fluid (CSF), these calibration methods rely on necessary corrections, when used in
heterogeneously composed voxels2,3. In this study an adopted method for heterogeneous voxel composition correction for the internal creatine reference standard is proposed and concentrations obtained with IWR and tCr as reference were compared.
Subjects and Methods: The measurement settings were as follows: 3 T (Philips), PRESS TE/TR = 25/1600 ms, 32 averages, VAPOR, IVS, 16 ml voxel placed in the right dorsolateral prefrontal cortex, 18 healthy subjects. For IWR additional 8 interleaved scans were acquired without
water suppression. Spectra were fitted using LC Model. Tissue segmentation was performed on a T1-weigthed image using SPM8. For the calculation ofmoles of metabolite m per volume of brain tissue cm [mM/l], the resonance areas Sm were referenced to tCr (cm, tCr) or to internal water (cm, tH2O). The formulas used for both references are shown below. Only the fitted resonance area of tCr and the internal water in these ratios were corrected with the relaxation attenuation factors R [Table 1]3,4,5.
Results: The metabolites could be fitted with low mean Crame´r-Rao lower bounds over all subjects (tCr:3.5, tNAA:3.5, mI:5, Glu + Gln:7, tCho:6.5). With the described corrections, similar mean concentrations for all metabolites were obtained (Fig. 1a, b). Without the
partial volume correction for the IWR (Fig. 1c), or by just assuming a pure GM tCr concentration (Fig. 1d), this could not have been realized, underlining the necessity of these corrections. Using the IWR we obtained a mean tCr concentration of 7.8 ± 0.6 [mM/l], which is in good agreement with the 7.7 [mM/l] predicted with the values given above and the average voxel composition (fGM,mean = 0.54, fWM,mean = 0.36). The correlation of the metabolite
ratios to tCr with the ratios to water (Fig. 2) is small and decreased with the applied corrections.
Discussion/Conclusion: In conclusion, the mean metabolite concentrations are only in agreement for both reference standards IWR and tCr, when the tissue composition corrections are applied. The decrease in correlation indicates that in this study the corrections did
decrease the agreement between two reference standards within the same subjects. This is potentially explained by inaccurately determined voxel compositions or potential individual deviations from the assumed relaxation properties and reference concentrations.
