Simultaneous mapping of metabolites and individual macromolecular components via ultra‐short acquisition delay 1H MRSI in the brain at 7T

    loading  Checking for direct PDF access through Ovid


Proton MR spectroscopic imaging (1H‐MRSI) is a powerful, noninvasive technique that provides valuable insights into brain metabolism. 1H‐MRSI benefits from the increased chemical shift dispersion and signal‐to‐noise ratio (SNR) at ultra‐high magnetic field strengths ( ≥ 7T). To take full advantage of the ultra‐high field, ultra‐short acquisition delay (TE*) free induction decay (FID)‐MRSI sequences recently have been developed 1. With FID‐MRSI, relaxation‐related SNR losses are minimized and J‐coupling evolution is eliminated, thus allowing the quantification of more metabolites compared to what is available at lower fields ( ≤ 3T) 5. However, MR spectra measured with such sequences contain prominent high‐molecular‐weight macromolecules (MMs) signals superimposed on the signal of low‐molecular‐weight metabolites. These MM contributions are particularly strong due to their short T1 and T2 relaxation times.
The presence of such broad MM signals in proton spectra of the brain already was described in the early 1990s 6. Behar et al. assigned these MM resonances to cytosolic proteins, mostly to the methyl and methylene groups of protein amino acids. The signal of MMs in the range from 0 to 4.7 parts per million (ppm) consists of 10 individual peaks: 0.90 ppm (MM1); 1.21 ppm (MM2); 1.43 ppm (MM3); 1.67 ppm (MM4); 2.04 ppm (MM5); 2.26 ppm (MM6); 2.99 ppm (MM7); 3.21 ppm (MM8); 3.8 to 4.0 ppm (MM9); and 4.3 ppm (MM10) 7.
Because MM signals are very strong in short‐TE*/TE spectra, the omission of MM contributions in the fitting routine may yield substantial errors in the quantified metabolite levels 7. Moreover, the quantification of these strong MM signals, per se, may provide valuable information. MM concentrations were found to be age‐ and region‐dependent in the healthy brain 8. In addition, several studies have shown MM levels to be altered in various diseases 11. This makes them potentially valuable for clinical studies.
Many possibilities to properly handle MMs during quantification were proposed 7. In most cases, MMs only are accounted for to improve metabolite quantification. At lower field strengths ( ≤ 3T) and shorter TEs, in which MM resonances are mere bumps rather than distinct peaks, a mathematical estimation using a spline baseline is adequate 14. At field strengths > 3T, such a simulation of the MM background usually is insufficient. Moreover, there has been an increased interest in MM quantification over the last couple of years. Previously published papers modeled the MMs at ultra‐high field with a single measured MM spectrum (typically a metabolite‐nulled MM spectrum acquired with inversion recovery methods) 10. These measured MM spectra are a well‐established MM model for metabolite quantification but do not allow for quantification of the individual MM peaks. Consequently, the quantification of metabolites with such a model only will fail when individual MM resonances are pathologically altered 13. Multiple molecules and chemical groups typically contribute to even single MM peaks, which impedes any effort for quantum‐mechanical simulations as applied for metabolite signals. Hence, we aimed to derive the individual MM basis spectra by parameterization of in vivo metabolite‐nulled spectra 13. If the MM basis spectra are combined with metabolite basis spectra in one basis set, MMs and metabolites simultaneously can be quantified from FID‐MRSI spectra.
The main goal of this work was to simultaneously map individual macromolecule components and metabolite levels in a healthy human brain at 7T. The information about the individual MM resonances may provide a better understanding of MM pathological changes in a diseased brain.

Related Topics

    loading  Loading Related Articles