Purpose: To evaluate whether vitamin C is detectable using clinical routine MRS combined with LCModel analysis and verify the reliability of its estimated concentration.
Materials and methods: In order to determine the reliability of the estimated ascorbic acid (Asc) concentrations, we analyzed 76 in vivo single voxel spectra (SVS) acquired on a 3T scanner (point-resolved spectroscopy with TE = 30 ms, TR = 3000 ms, voxel size = 8 mL, without spectral editing) from different regions within the brain using LCModel. In addition to this we simulated multiple concentration levels by adding adapted Asc spectra to the in vivo data.
Results: Asc was successfully detected in 71 of 76 in vivo spectra with a comparable concentration ratio to that of myo-inositol. Furthermore, we observed good linearity between added Asc concentrations and the LCModel estimates within the simulations having correlation coefficients larger than 0.985 at three different linewidth cases.
Conclusion: This study demonstrates the ability to detect vitamin C in the human brain under common clinical magnetic resonance spectroscopy (MRS) standards in combination with LCModel. Furthermore, it supports the need to include Asc in the standard MRS analysis.
(c) 2008 Wiley-Liss, Inc.