Comparison of MRS acquisition methods for separation of overlapping signals at 3 T

Background

Proton magnetic resonance spectroscopy (MRS) can be used to quantify multiple neurometabolites. However, due to the difficulty of separating overlapping signals at the commonly used field strength of 3 T, the quantified values are often composites of metabolically related chemicals. This can complicate interpretation and mask effects of interest. Therefore, it is important to determine the ability to accurately separate these signals at 3 T. Data acquired at 7 T can provide a benchmark, as higher field strength facilitates spectral resolution and reduces the signal overlap.

New Methods

This study assessed the ability of multiple 3 T MRS sequences to separate the commonly acquired neurochemicals (Glutamate (Glu) and Glutamine (Gln); N-Acetyl aspartate (NAA) and N-acetylaspartylglutamate (NAAG); Creatine (Cr) and Phosphocreatine (PCr); Choline (Cho), Phosphocholine (PC) and Glycerophosphocholine (GPC)). We compared metabolites quantified at 3 T from 6 acquisitions (PRESS, TE= 20, 30, 40, 80 ms, semi-LASER, TE=28 ms and STEAM TE=6 ms) with those quantified at 7 T using STEAM (TE=8 ms).

Results

Sequences with short echo times (STEAM-6, PRESS-20) generally performed better at separating most metabolites when using correlative and difference analyses with 7 T reference data. The exceptions were NAAG, which was best quantified with PRESS-80, and Cr and PCr, which were not well separated by any sequence.

Comparison with existing methods and conclusion

When wanting to specifically separate composite metabolite signals using single voxel MRS, shorter echo times generally perform better. Researchers should be mindful of the effects of acquisition parameters on the metabolite measures.