Sapporo Medical Journal

Abstract : Introduction: Glutamine (Gln) and glutamate (Glu) signals are highly overlapped in the region of 2-2.5 ppm in human brain due to the resemblance of their spin systems, making it difficult to separate these metabolites using short-TE PRESS or STEAM at 1.5 T. These metabolites are highly important markers for many diseases, and accurately estimation of them is important. Methods: This study reports the detection of Glu and Gln at 3 T, and immune to the frequency drift. The echo times were optimized, with numerical analysis of the filtering performance, as {TE(1), TE(2), TE(3)} = {23, 74, 18} ms. The amplitude ratios of the filtered Glu and Gln multiplets with respect to 90°-acquisition were 28% and 60%, respectively. Results: Calculation indicated that the amplitude of filtered Gln peak is two-fold compared to STEAM scheme, regardless of the relaxation effects. Glu and Gln are detected at 2.29 ppm and 2.39 ppm simultaneously, without substantial interferences from the neighboring resonances in vivo. The filtered sum spectra with and without Gln in LCModel analyses, using density-matrix calculated spectra as basic functions. The concentration of Glu and Gln in hippocampus of healthy adults was estimated to be 2.8±0.7, and 10.5 ±0.6 mM respectively, and NAA of 9.7±0.6 mM, with reference to creatine at 8 mM. The fit standard deviations are10± 2%, 6±2 % and 2±1% respectively. Conclusion: The presented result shows the fully resolved and Glu and Gln which helps in accurate es