INTRODUCTION

Because of scalar couplings, many metabolites are best detected at a predefined TE. Except at very long TEs, the presence of varying macromolecule signals may interfere with measuring metabolite T₁. High field MRS provides increased sensitivity and spectral resolution compared to lower field strengths. However, metabolite T₁ relaxation times become longer at high field strengths, which often makes it necessary to characterize metabolite T₁ values to obtain accurate quantification of metabolite concentrations and/or optimize the TR of a pulse sequence for optimal signal-to-noise performance. A previously described spectral editing technique¹ can simultaneously measure glutamate (Glu), glutamine (Gln), γ-aminobutyric acid (GABA), and glutathione (GSH) at 7 T. The editing RF pulse was set to OFF, ON at 1.89 ppm, and ON at 2.12 ppm. By numerical optimization of sequence timing in the presence of an editing pulse, all four metabolites form relatively intense pseudo singlets at TE = 56 ms with maximized peak amplitudes and minimized peak linewidths in one of the three interleaved spectra. In this work, we propose to combine the spectral editing technique with inversion recovery (IR) to measure metabolite T₁ relaxation times in the presence of substantial macromolecule baseline signals.

METHODS

To incorporate IR, a 180° hyperbolic secant pulse was placed before the water suppression pulses in the spectral editing pulse sequence¹ (Figure 1). Three IR settings were used: no inversion, TI = 2150 ms, and TI = 1600 ms. With average T₁ = ~430 ms, 99% of macromolecules were recovered at TI = 2150 ms and 95% were recovered at TI = 1600 ms². Therefore, macromolecule signals were minimally affected by the IR pulse. In contrast, with T₁ = ~1600 ms, metabolite signals only recovered by 62% at TI = 2150 ms and 41% at TI = 1600 ms. Hence, the metabolite signals were significantly modulated by the TI values.

Five healthy volunteers were recruited and consented for the study in accordance with procedures approved by our institutional review board. Experiments were performed on a Siemens Magnetom 7 T scanner. MRS data were collected using the proposed pulse sequence from a 2 × 2 × 2 cm³ voxel in the pregenual anterior cingulate cortex (pgACC) of the volunteers. To compute within-subject coefficient of variation (CV) values on each subject, the same MRS measurement was performed three times in the same exam using the same voxel prescription. Interleaved data acquisitions were performed using nine different parameter settings which were a combination of the three spectral editing settings and three IR settings. The spectra corresponding to the nine parameter settings were simultaneously fitted in the range of 1.8 – 3.7 ppm by a linear combination of numerically computed basis functions of acetate, N-acetylaspartate (NAA), N-acetylaspartylglutamate (NAAG), GABA, Glu, Gln, GSH, aspartate, total creatine (tCr), total choline (tCho), myo-inositol (mIns), taurine, scyllo-inositol, and glycine, as well as cubic spline baselines which approximate the macromolecule signals. Each basis function consisted of three FIDs which correspond to the three different spectral editing settings. The unknown variables in the fitting process were concentrations, T₁s, frequency shifts, linewidths, and lineshape of the metabolites, as well as zero-order phases of the nine spectra and control points of the cubic spline baselines. Because of the low concentrations of Gln and GABA, it was difficult to measure their T₁ values. Hence, the T₁ values of Gln and GABA were set to be the same as that of Glu. Due to the long TIs, macromolecule signals were barely affected by the IR pulse. Hence, baselines were not changed for the three different IR settings.

RESULTS

Reconstructed spectra and corresponding fits from one healthy volunteer are displayed in Figure 2. The spectra showed progressively reduced metabolite signals as TI decreased, whereas the same baselines were used for the three different IR settings. The in vivo spectra were fitted very well and the fit residuals were small. Metabolite concentrations (/[tCr]) and T₁ relaxation times (n=5) are given in Table 1. The metabolite concentrations and T₁ values agree well with previous studies^{1, 3}.

DISCUSSION AND CONCLUSION

In this study, the TI values were chosen in such a way that the metabolite signals were changed by as much as 60% while the macromolecule signals were essentially intact. This avoided the complex T₁ relaxation effects of different components of macromolecules, and thus simplified data acquisition and post-processing by using the same baselines for data acquired at different TIs. If shorter TI values were used, macromolecule signals could not fully recover and the T₁ distribution of macromolecules needs to be determined as in a previous study⁴, which would require much longer scan time and more time-consuming post-processing. Despite the reduced sensitivity for measuring metabolite T₁ values due to using only relatively long TI values, high precision for metabolite concentrations and T₁s was still achieved. In summary, this work combined inversion recovery with a previously described spectral editing technique to measure metabolite T₁ relaxation times at TE = 56 ms in the presence of substantial macromolecule baseline signals. Metabolite concentrations and T₁ relaxation times were measured with high precision.

Acknowledgements

This work was supported by the intramural programs of the NIH.