Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2012 May 22.
Published in final edited form as: NMR Biomed. 2011 Jul 27;25(2):332–339. doi: 10.1002/nbm.1754

Localized 1H NMR Spectroscopy in Different Regions of Human Brain In Vivo at 7 T: T2 Relaxation Times and Concentrations of Cerebral Metabolites

Małgorzata Marjańska a, Edward J Auerbach a, Romain Valabrégue b, Pierre-François Van de Moortele a, Gregor Adriany a, Michael Garwood a
PMCID: PMC3357544  NIHMSID: NIHMS372278  PMID: 21796710

Abstract

At the high field strength of 7 T, in vivo spectra of the human brain with exceptional spectral quality sufficient to quantify sixteen metabolites have previously been obtained only in the occipital lobe. However, neurochemical abnormalities associated with many brain disorders are expected to occur in brain structures other than the occipital lobe. The purpose of the present study was to obtain high quality spectra from various brain regions at 7 T and to quantify the concentrations of different metabolites. To obtain concentrations of metabolites within four different regions of the brain such as occipital lobe, motor cortex, basal ganglia, and cerebellum, the T2 relaxation times of the singlets and J-coupled metabolites in these regions were measured for the first time at 7 T. Our results demonstrate that high quality, quantifiable spectra can be obtained in regions other than the occipital lobe at 7 T utilizing a 16-channel transceiver coil and B1+ shimming.

Keywords: occipital lobe, motor cortex, basal ganglia, cerebellum, LCModel

Introduction

Proton Magnetic Resonance Spectroscopy (1H MRS) provides a noninvasive way to investigate in vivo neurochemical abnormalities of many brain disorders. Each observable metabolite can potentially provide unique information about brain biochemistry and be a biomarker for brain disorders to facilitate diagnosis and treatment. MRS data can be obtained using either chemical shift imaging (CSI) or single-voxel localization. In general, CSI data are usually acquired within a slice and allow the quantification of a limited number of metabolites (up to eight) (1–4). Three-dimensional (3D) CSI acquisitions at 3 T enabled the detection of five metabolites (5). Previous 3D CSI acquisitions performed at 7 T had a limited spatial coverage and five metabolites were detected without the ability to distinguish between glutamate and glutamine (6, 7). On the other hand, single-voxel data can be potentially obtained from any brain region, and up to sixteen metabolites can be quantified (8, 9). At the high field strength of 7 T, exceptional spectral quality that allows for the quantification of sixteen metabolites has been previously obtained in occipital lobe (8–12). However, neurochemical abnormalities for many brain disorders are expected to occur in brain regions other than the occipital lobe.

Previous single-voxel 7 T studies that measured sixteen different neurochemicals used localization techniques with short echo time (6 ms in STEAM (8) and SPECIAL (9)). The absolute quantification when using these techniques is straightforward since there is no need to consider J evolution and T2 relaxation. An alternative single-voxel technique is a fully-adiabatic spin-echo localization by adiabatic selective refocusing (LASER) (13) sequence, which is a single-shot sequence that generates very sharp localization edges without the need for outer-volume suppression. The adiabatic full passage (AFP) pulses used in LASER are robust to radio frequency (RF) field changes, which makes them very advantageous at high fields where B1 inhomogeneity is problematic. The available peak RF power also does not limit the bandwidth which minimizes the voxel displacement. However, LASER tends to require a relatively long echo time. To shorten the echo time, the adiabatic half passage (AHP) pulse and one pair of AFP pulses can be replaced with a single slice selective pulse (14, 15). Even with that replacement, the echo time is much longer than 6 ms used in the STEAM or SPECIAL sequences. To obtain absolute concentrations of metabolites from spectra obtained at longer echo times, the knowledge of both the J-modulation and T2 relaxation times is required. The T2 relaxation times at 7 T have been reported previously for the methyl protons of N-acetylaspartate (NAA) and the methyl protons of total creatine (creatine + phosphocreatine, tCr) (10, 16).

The measurement of T2 for J-coupled metabolites is more difficult as their signal intensity is decreased with echo time and their spectral pattern changes due to J-evolution. For reliable T2 measurements, the signal changes due to J-modulation need to be taken into account by simulating the spectral pattern at each time. Additional complications arise from the overlap of resonances from different molecules and macromolecule contributions. Recently, T2 relaxation times of J-coupled cerebral metabolites in rat at 9.4 T were obtained using simulated echo time (TE) specific basis sets and LCModel analysis (17). Excellent fits were obtained when each metabolite was considered as one component except for NAA and tCr which were split into NAA methyl and aspartate moieties, and CH3 and CH2 groups, respectively suggesting that correct prior knowledge was used to perform the analysis.

The aim of the present study was to investigate the efficacy of using the LASER sequence to obtain high quality 7 T spectra from brain regions other than occipital lobe, to measure T2 relaxation times in different brain regions of the singlets and J-coupled metabolites, and to quantify concentrations of metabolites in different brain regions. The obtained results show that high quality, quantifiable spectra can be obtained in vivo at 7 T in regions of the brain other than the occipital lobe.

Experimental

Subjects

Healthy, normal volunteers (n = 23, 12 M, 11 F, age: 23 ± 4 years) were studied after giving informed consent according to the procedures approved by the Institutional Review Board of the University of Minnesota Medical School.

MR Acquisition

In vivo data were obtained on a 7-T, 90-cm horizontal bore magnet (Magnex Scientific Inc., Oxford, UK) interfaced with a Siemens TIM console (Siemens, Erlangen, Germany). The magnet was equipped with a gradient coil capable of reaching 40 mT/m in 200μs. A 16-channel transmission line head array RF coil (18) was used to transmit and receive. The transmit phase of each coil channel was controlled with independent 1 kW RF amplifiers (CPC, Brentwood, NY, USA). A RF power monitoring system measured forward and reflected power for each channel to ensure that local specific absorption rate remained below 3 W/kg.

Turbo spin echo (TSE) images (repetition time, TR = 3 s, echo time, TE = 95 ms, field-of-view, FOV = 26 × 26 cm2, matrix = 256 × 256, slice thickness = 1.5 mm, 66 slices, acquisition time = 4 min) were acquired with a 22.5° RF transmit phase difference between adjacent elements (standard transmit phases) to select the volumes of interest (VOI) in occipital lobe (OC, 2.7 × 2.7 × 2.7 cm3), motor cortex (MC, 2 × 2 × 2 cm3), basal ganglia (BG, 1.5 × 4 × 1.5 cm3), and cerebellum (CR, 2.5 × 2.5 × 2.5 cm3). The occipital VOI was positioned symmetrically within the occipital lobes of both hemispheres. The motor cortex VOI was centered on the hand knob, which corresponds to the primary motor cortex of the upper limb representation (19). The basal ganglia VOI was positioned in the lentiform nucleus and therefore included the putamen and globus pallidus. The cerebellar VOI was placed symmetrically about the midline and mainly included the cerebellar vermis.

Low flip angle gradient echo images were acquired for B1+ shimming by pulsing RF power through one channel at a time and receiving the signal with all 16 channels (TR = 60 ms, TE = 3.6 ms, FOV = 168 × 256 cm2, nominal flip angle = 15°, slice thickness = 4 mm, total acquisition time = 2.5 min). For each voxel, the B1+ phase was optimized based on a previously published algorithm (20). A set of transmit phases determined by B1+ shimming improved transmit efficiency by a factor of 2 ± 0.5 (mean and standard deviation) in the occipital lobe, 1.31 ± 0.09 in the motor cortex, 1.35 ± 0.04 in the basal ganglia, and 3 ± 1 in the cerebellum when compared to standard transmit phases.

All spectra were acquired using a LASER sequence in which the AHP pulse and two AFP pulses were replaced with a slice-selective 90° Hamming-filtered sinc pulse (14, 15), a pulse length of 1.52 ms and a bandwidth of 5.84 kHz. The two other dimensions were selected with a pair of hyperbolic secant pulses, HS1 (13, 21, 22), with a pulse length Tp of 5.12 ms and a bandwidth of 3.92 kHz. Each voxel measurement began with a calibration of the localized B1+ field strength and an adjustment of the first- and second-order shims using FAST(EST)MAP (23, 24). The water suppression module consisted of variable power and optimized relaxation delays (VAPOR) and outer volume suppression (OVS) (25) in the direction of slice excitation only. These modules were adapted for the human 7 T system and incorporated prior to the excitation pulse in LASER. The repetition time was 4.5 s.

Each free induction decay (FID) was acquired with 2048 complex points using a spectral width of 4 kHz. FIDs were stored separately in memory and then frequency and phase corrected based on either the NAA signal at 2.01 ppm or the tCr signal at 3.03 ppm prior to summation.

For the T2 measurement, the echo time was extended by adding delays around the last AFP pulse in the sequence, and the spectra were collected at six echo times: 35, 70, 105, 140, 175, 210 ms. For all echo times, 64 averages were acquired. Macromolecular (MM) spectra were acquired using the inversion recovery technique (26) (TR = 2 s, TIR = 900 ms, number of scans (NS) = 1920 (TE = 35 ms), NS = 1664 (TE = 70), NS = 1664 (TE = 105 ms), and NS = 512 (TE = 140 ms)) in the occipital lobe of five subjects for all TE’s except 175 and 210 ms where the MM signal was very weak. For assessing regional differences in metabolite concentrations, spectra were acquired with an echo time of 35 ms and 128 averages.

Spectral Fitting

The acquired spectra were analyzed using LCModel 6.1-4A (27, 28) (Stephen Provencher, Inc., Oakville, Ontario, Canada), which calculates the best fit of the experimental spectrum as a linear combination of model spectra. The basis set for LCModel was generated using home-written programs based on the density matrix formalism (29) in Matlab (The MathWorks, Inc., Natick, MA, USA) using the known chemical shifts and J-couplings (30, 31). In addition, a small signal was added at 0 ppm in all basis spectra for automatic frequency referencing. The simulated spectra of the following twenty one metabolites were included in the basis set for LCModel: alanine (Ala), ascorbate (Asc), aspartate (Asp), creatine (Cr), γ-aminobutyric acid (GABA), glucose (Glc), glutamine (Gln), glutamate (Glu), glycerophosphorylcholine (GPC), glycine (Gly), glutathione (GSH), lactate (Lac), myo-inositol (mIns), NAA, N-acetylaspartylglutamate (NAAG), phosphocreatine (PCr), phosphorylcholine (PCho), phosphorylethanolamine (PE), scyllo-inositol (sIns), taurine (Tau) and threonine (Thr). Three metabolites, NAA, Cr, and PCr, were separated into different moieties, and independent spectra for the singlet (CH3 group, sNAA) and the multiplet (CH2 group, mNAA) of NAA and the singlet of the CH3 and CH2 groups of tCr were included in the basis set. No baseline correction, zero-filling or line broadening were applied to the in vivo data prior to the analysis.

The experiment MM spectrum (average from five subjects) from occipital region at each TE was included in the basis set, except for TE of 175 and 210 ms. The methylene protons peak of tCr at 3.93 ppm was present in the MM spectra due to a shorter T1 relaxation time (26) and was removed in the time domain using the HSVD Lanczos algorithm from MRUI 99.2b (32). The LCModel fitting was performed over the spectral range from 0.5 to 4.2 ppm.

Signal to Noise Ratio Calculation

Signal to noise ratios (SNR) for various metabolites were calculated using prior knowledge and taking the ratio of the maximum signal for the particular metabolite minus baseline over root mean square of the noise measured between 0.2 and 0.4 ppm. No line-broadening was applied.

T2 fitting

The T2 values of various metabolites were determined by fitting the concentrations obtained from an LCModel analysis (without water scaling) using a two parameter mono-exponential decay function with a non-linear least square algorithm (nlinfit) in Matlab. The goodness of fit was evaluated using corrcoef function in Matlab which returns p values for testing the hypothesis of no correlation. If the p value is small, then the correlation is significant.

Quantification

Quantification was obtained using the unsuppressed water signal obtained from the same voxel (33). Concentrations were corrected for cerebrospinal fluid (CSF) content. The tissue composition was measured using TSE images, which were segmented into gray matter (GM), white matter (WM) and CSF content using the segmentation function of SPM8 software (www.fil.ion.ucl.ac.uk/spm). The fractional volumes of GM, WM and CSF were obtained by averaging the gray/white and CSF segmented images over the spectroscopic voxel. The relative densities of MR-visible water for GM, WM, and CSF were assumed to be 0.78, 0.65, 0.97 (34), respectively. The T1 and T2 relaxation times of water used in the calculation of the attenuation factors were taken from published reports (T1(GM) = 2130 ms, T1(WM)= 1220 ms, T1(CSF) = 4425 ms (35); T2(GM) = 50 ms, T2(WM) = 55 ms, T2(CSF) = 141 ms (36)). The water attenuation was computed using the fractional volume of each compartment. For the basal ganglia voxel, the values of 1523 ms for T1 and 41.2 ms for T2 of water were used based on the previously reported T1 values of water in putamen and globus pallidus (35) and T2 measured in this study (Table 1), respectively. For the cerebellum, the value of 48 ms was used for T2 of water (Table 1). The T1 relaxation for the metabolites was not taken into consideration as an attenuation factor since this would be very small due to the use of repetition time of 4.5 s. The T2 of the metabolites obtained in this study were used. The T2’s of the metabolites which we were unable to measure were assumed to be similar to that of mIns (37), while Gln and NAAG were considered to have similar T2 values as Glu and the singlet of NAA, respectively.

Table 1.

T2 values (mean ± SD) of water and metabolites measured at 7 T from three subjects per voxel (10 subjects in total). For each substance (water or metabolites) means with different letters in the superscript differ at 0.05 significance level. For all singlet resonances, fits with R2 ≥ 0.97 were obtained except for sIns in BG, where R2 = 0.85. For all multiplet resonances, fits with R2 ≥ 0.8 were obtained except for mNAA in CR, where R2 = 0.54.

Voxel OC MC BG CR
Compound Group T2 (ms) R2 T2 (ms) R2 T2 (ms) R2 T2 (ms) R2
water 47 ± 1b 0.9991 47 ± 1b 0.9996 41.2 ± 0.8a 0.99993 48 ± 3b 0.9997
singlets NAA 2CH3 132 ± 6a 0.994 168 ± 6b 0.996 130 ± 11a 0.97 191 ± 7c 0.998
tCr N(CH3) 95 ± 3a, b 0.9997 113 ± 2b, c 0.9988 90 ± 11a 0.996 131 ± 8c 0.9991
tCr 2CH2 84 ± 2a 0.9985 108 ± 5b 0.995 81 ± 15a 0.994 102 ± 3a, b 0.998
tCho entire molecule 152 ± 3a 0.992 139 ± 9a, b 0.997 121 ± 5b 0.984 200 ± 17c 0.996
sIns 96 ± 8a 0.95 112 ± 4a 0.96 80 ± 20a 0.85 130 ± 20a 0.87
J-coupled NAA 3CH2 90 ± 11b 0.80 110 ± 30b 0.80 69 ± 12b 0.85 170 ± 12a 0.536*
Glu entire molecule 93 ± 4b 0.95 98 ± 4b 0.985 88 ± 10b 0.96 139 ± 8a 0.93
GSH entire molecule 61 ± 3a 0.88 97 ± 8b 0.74# 80 ± 10a, b 0.908
mIns entire molecule 95 ± 2b 0.982 100 ± 15b 0.988 87 ± 6b 0.988 160 ± 20a 0.985
Tau entire molecule 93 ± 7a, b 0.80 90 ± 16a, b 0.8 85 ± 10b 0.89 120 ± 20a 0.83
Open in a new tab

Means with a different letters were significantly different with the significance level of 0.05. Means with the same letter were not significantly different.

All p values were below 0.04 for each subject, each metabolite and each region except for (*) mNAA in CR where p values were 0.095, 0.099, 0.099, and for (#) GSH in MC where for one subject p value was 0.1.

The criteria for selecting reliable metabolite concentrations were based on the Cramér-Rao lower bounds (CRLB), which are estimates of the %SD of the fit for each metabolite (27). Only results with a CRLB ≤ 50% were included in the analysis. Concentrations with CRLB > 50% were classified as not detected. If the covariance between two metabolites was consistently high (correlation coefficient < −0.5), such as was found for Cr and PCr, their sum was reported.

Statistical Analysis

Statistical analysis was conducted using SAS Software for Windows (version 9.1, SAS Institute, Cary, NC). One-way analysis of variance (ANOVA) with a Tukey post hoc test was used to compare the T2 relaxation times and the concentrations at each location for each metabolite. This is appropriate because a different set of subjects was used to estimate concentrations and relaxation times and thus the four regions represent four independent groups. A two-sample paired t-test was used to compare the T2 relaxation times between sNAA and mNAA, and between tCr at 3.03 ppm and 3.93 ppm. The paired t-test was used to account for the correlation between two moieties of the same metabolite from the same subject.

Results and Discussion

Spectral Quality

In Figure 1, boxes drawn on the T2-weighted images of a human brain obtained at 7 T show the position and size of the localized volumes from where the spectra were acquired. Voxels of different sizes were placed over the occipital lobes in both hemispheres, the primary motor cortex of the upper limb representation, the basal ganglia in lentiform nucleus and the cerebellum. The spectra shown in Figure 1 are representative of the quality consistently obtained using a modified LASER sequence in this study.

Figure 1.

Figure 1

Open in a new tab

LCModel quantification of the representative 1H NMR spectra obtained at 7 T with a modified LASER sequence from the four following regions of the human brain: (a) a 19.7 mL voxel placed in the occipital lobe (OC), (b) an 8 mL voxel placed in the motor cortex (MC), (c) a 9 mL voxel placed in the basal ganglia (BG), and (d) a 15.6 mL voxel placed in the cerebellum. In all spectra, TR = 4.5 s, TE = 35 ms, NS = 128, and no line-broadening was applied. The location and size of the voxels are shown on T2-weighted images.

The spectra were of very high quality without any contamination from signals outside the voxel such as the lipid signal which suggested that the modified LASER sequence was very efficient in suppressing signals from outside of the voxels. Similar linewidths of water (13 ± 1 Hz in OC, 11 ± 2 Hz in MC, 14 ± 1 Hz in CR, and 21 ± 4 Hz in BG) and tCr (12 ± 2 Hz in OC, 12 ± 1 Hz in MC, 11 ± 1 Hz in CR, and 22 ± 2 Hz in BG) were obtained in all regions except for the basal ganglia. The linewidths of water and total creatine were very similar to each other in all voxels. In the basal ganglia, it turned out to be more difficult than in other regions to achieve good field homogeneity by localized shimming, possibly due to the heterogeneity of the tissue included in the voxel. The variation of the signal-to-noise ratio in these four spectra is mainly due to the differences in the voxel size and linewidths. Additionally, the residual of the basal ganglia spectrum has uneven noise due to poorly spoiled water signal. The spectral quality was degraded in the basal ganglia region due to a 10 Hz linewidth difference between the spectra obtained in the basal ganglia and other regions. Excellent water suppression was also consistently achieved, as illustrated by the small water residual observed in all subjects (data not shown) and by the flat baseline (Figure 1).

It was possible to obtain these spectra due to our ability to control the amount of available B1+ over a particular region of interest using local B1+ shimming (20). Figure 2 demonstrates the gain in transmit B1 obtained after local B1+ shimming. With the standard phase arrangement, the most efficient transmission was at the center of the brain, but even for the two voxels located relatively close to the center of the brain, motor cortex and basal ganglia, the gain in the transmit efficiency was significant (~1.3). The biggest gain in transmit efficiency was observed for the voxels located in the periphery such as in the cerebellum and occipital lobe.

Figure 2.

Figure 2

Open in a new tab

The relative transmit efficiency before and after local B1+ phase shimming, optimized for the following regions of interest: (a) OC, (b) MC, (c) BG, and (d) CR. The average gain and standard deviation (n = 5) in the transmit efficiency is listed next to each image.

T2 Relaxation Times

Figure 3 shows the quality of representative spectra obtained from the occipital lobe at different echo times from one volunteer. The singlet resonances become smaller with increasing echo time while the multiplet resonances undergo J modulation. The macromolecule resonances (especially visible between 0.5 and 1.9 ppm) also undergo J evolution, which can be clearly seen from the inversion of the macromolecular resonances in the spectrum acquired at an echo time of 105 ms. Due to their short T2 and scalar coupling, macromolecular resonances decrease as a function of echo time, and after 140 ms, these signals are below the thermal noise level. As a result, long echo time spectra are characterized by a flat baseline in the region between 0.5 and 1.9 ppm. In acquiring the macromolecular spectra, the decreased relative contribution of macromolecular signals to the overall spectrum allowed the number of averages to be reduced as the echo time increased without having an affect on the quality of the fits.

Figure 3.

Figure 3

Open in a new tab

Representative 1H NMR spectra obtained at 7 T with a modified LASER sequence from the 19.7 mL voxel placed in the human occipital lobe. For the spectra, TR = 4.5 s, NS = 64, and no line-broadening was applied.

The SNR for tCr at 3.03 ppm as a reference and J-coupled metabolites are reported in Table 2 for the data displayed in Figure 3. The SNR for J-coupled metabolites reflects the effect of both T2 relaxation and J evolution. The lowest SNR values of 1.95 for GSH and 3.8 for Tau demonstrate that these metabolites can be readily detected at all echo times.

Table 2.

SNR values of metabolites measured from spectra in Figure 3.
SNR
TE (ms) tCr at 3.03 ppm mNAA Glu mIns GSH Tau
35 292.9 51.0 82.3 78.5 37.2 23.2
70 210.9 39.8 57.8 72.6 18.9 7.8
105 187.8 30.5 80.3 76.2 17.4 8.4
140 115.9 23.6 49.9 24.5 15.5 7.6
175 97.9 51.1 22.9 39.6 1.9 7.8
210 60.3 17.9 18.0 31.1 4.8 3.8
Open in a new tab

The mean values and standard deviations (SDs) for the T2 values of water and metabolites measured at 7 T in four brain regions are listed in Table 1. The mean R2 values, which indicate the accuracy of the exponential fits, are also listed in Table 1. The relative SDs were typically 5% to 10%, but were larger for the aspartyl resonances of NAA (mNAA) in all the brain regions, for all metabolites in the basal ganglia, for sIns, Tau and GSH resonances in the cerebellum, and for the mIns and Tau resonances in the motor cortex. Additionally, the goodness of fit was evaluated and p values are reported in Table 1. All p values were below 0.04 except for mNAA in cerebellum and GSH in motor cortex which suggested that good quality fits were obtained for all metabolites in all regions except for those four cases.

The T2 relaxation times of water were the same in all voxels except for the basal ganglia where it was significantly shorter. The T2 value obtained for water in the relatively large occipital lobe voxel containing GM, WM, and CSF was within the standard deviation of the T2 previously measured in the GM and slightly shorter than average value plus the standard deviation of the T2 previously measured in the WM with LASER at 7 T (36).

The T2 values for sNAA ranged from 130 to 191 ms in different brain regions. In the occipital lobe and the basal ganglia, T2’s of sNAA were the same and were significantly shorter than in the motor cortex and the cerebellum. In this study, the T2 value obtained in the occipital lobe was in agreement with the lower end of the published T2 range (10, 16).

The T2 values for the methyl protons of tCr ranged from 90 to 131 ms in different brain regions. They were the same in the occipital lobe and the basal ganglia, but significantly longer in the cerebellum. In the motor cortex, the T2 value for the tCr peak at 3.03 ppm was not significantly different from that found in the cerebellum and the occipital lobe, but it was significantly longer than that found in the basal ganglia. The measured T2 relaxation time for tCr (CH3 peak) in the occipital lobe, 95 ± 3 ms, was within a standard deviation of previously measured values at 7 T (10, 16).

Other relaxation times reported in Table 1 for both singlets and J-coupled metabolites were measured for the first time at 7 T in human brain. The T2 value for mNAA was significantly shorter in the occipital lobe region (p = 0.006) than for sNAA while there was no significant difference between the T2 values of sNAA and mNAA in the basal ganglia and the cerebellum. In the motor cortex, a tendency for the T2 of mNAA to be shorter (p = 0.096) than the T2 of sNAA was observed. It has been previously reported that there is a significant difference between T2 relaxation times of sNAA and mNAA in the parieto-occipital lobe (38). In that study and this study, the signal of mNAA decayed faster than the sNAA signal. Additionally, a significant difference in the relaxation times for sNAA and mNAA has been reported in rat brain at 9.4 T (17).

The T2 values for the tCr peak at 3.93 ppm ranged from 81 to 108 ms and were significantly shorter than that found for the 3.03 ppm singlet in the occipital lobe (p = 0.0023) and in the cerebellum (p = 0.0164). The observed difference between the relaxation times of the tCr peaks at 3.03 ppm and 3.93 ppm has been reported previously (39). Additionally, the T2 of tCr (CH2 peak) was significantly longer in the motor cortex than in the occipital lobe and basal ganglia.

The T2 for the other singlets, tCho and sIns, ranged in values from 121 to 200 ms and 80 to 130 ms, respectively. The longest T2 value for tCho was obtained in the cerebellum, and the shortest was found in the basal ganglia. The T2 values for sIns and the methyl protons of tCr were very similar in all brain regions.

The T2 values obtained for the J-coupled metabolites measured in this study were generally shorter than the T2 values found for the singlets. In the occipital lobe, the T2’s of the J-coupled metabolites were very similar to the T2 of the methyl protons of tCr except for the T2 of GSH which was shorter. The longest T2’s of the J-coupled metabolites were measured in the cerebellum. The T2 of the J-coupled metabolites was not significantly different in the occipital lobe, motor cortex and basal ganglia voxels. Due to the difference in the quality of the spectra, it was not possible to obtain the T2 of GSH in the basal ganglia.

Quantification of Spectra at TE = 35 ms

The concentrations of metabolites (means and standard deviations) obtained from four brain locations after fitting spectra using the LCModel and correcting for T2 relaxation of water and metabolites, water tissue content, and CSF contribution, are listed in Table 3. Additionally, CRLBs are reported in Table 3 to reflect the estimated errors in quantification. The quality of the LCModel fits is shown in Figure 1 for each brain region. The baselines were flat with very small oscillations which most probably were due to differences arising from macromolecular spectrum, and the residuals were small without any noticeable resonances. The residual in the basal ganglia voxel was noisier than for other voxels due to the small size of the voxel and the broader linewidths than in other brain locations.

Table 3.

Average brain metabolite concentrations measured with a modified LASER sequence at 7 T (n = 5, 18 subjects, 6 subjects are the same as in Table 3) and quantified using LCModel, SD, and average CRLBs and SD of CRLB.

OC MC BG CR
Metabolite Conc. (μmol/g) SD (μmol/g) CRLB (%) SD (%) Conc. (μmol/g) SD (μmol/g) CRLB (%) SD (%) Conc. (μmol/g) SD (μmol/g) CRLB (%) SD (%) Conc. (μmol/g) SD (μmol/g) CRLB (%) SD (%)
Asc 0.4a 0.3 26 8 1.2a 0.5 19 7 1.5a 0.8 18 9 0.6a 0.2 28 13
Asp 2.9a 0.8 9 2 2.6a 0.3 14 2 1.2b 0.5 35 9 1.1b 0.4 27 14
GABA 1.5b 0.3 15 4 0.8b 0.6 29 9 1.7a 1.3 22 6 1.5b 0.5 18 6
Glu 9.6a 1.3 1.6 0.5 7.6b 0.9 2.4 0.5 6.7b 1 3.0 0.7 8.2a, b 0.6 2 0
Gln 1.5b 0.5 13 6 1.0b 0.3 29 7 1.6b 0.6 23 11 2.9a 0.7 7.2 0.8
Gly 0.7a 0.1 9 1 0.5b 0.1 20 5 -- -- -- -- 0.59a, b 0.07 12 2
GSH 1.1a 0.1 4.0 0.7 0.7a 0.1 7.6 0.5 1.1a 0.4 7 3 1.11a 0.06 4.4 0.5
mIns 6.4a, b 0.8 2 0 5.9a, b 0.3 2.6 0.5 5.1b 1.4 4.0 0.7 7.4a 0.8 2 0
sIns 0.4a 0.1 6 2 0.21a 0.06 13 4 0.5a 0.3 10 4 0.4a 0.2 7 3
NAA 12.6a 1.7 1 0 10.5a, b 1.2 1 0 7.8c 0.4 1.8 0.4 8.4b, c 1 1 0
NAAG 0.40a 0.2 18 6 1.6b 0.4 7 3 0.8a, b 1.1 21 18 1.0 a, b 0.2 6.0 0.7
PE 2.3a 0.4 7 1 1.1b 0.4 23 7 1.3b 0.5 19 3 1.3b 0.4 15 5
tCho 0.9c 0.1 2 0 1.32b, c 0.09 2.2 0.4 1.5b 0.4 3.2 0.4 2.2a 0.2 1 0
tCr 8.7a, b 1.1 1 0 7.4b 0.6 1.2 0.4 8b 2 2 0 10.2a 0.9 1 0
Lac + Thr 0.8a 0.2 10 4 0.9a 0.3 11 3 0.6a 0.1 20 10 1.0a 0.5 9 2
Glc + Tau 2.1b 0.3 7 2 1.7b 0.2 8 4 1.8b 0.6 9 1 3.5 a 0.4 4 1
Fractional tissue composition OC MC BG CR
CSF 0.05 ± 0.02 0.08 ± 0.02 0.001 ± 0.0009 0.07 ± 0.05
GM 0.51 ± 0.05 0.19 ± 0.03 0.25 ± 0.06 0.72 ± 0.14
WM 0.44 ± 0.05 0.73 ± 0.02 0.73 ± 0.05 0.19 ± 0.16
Open in a new tab

Means with a different letter were significantly different with the significance level of 0.05. Means with the same letter were not significantly different.

The concentrations of following metabolites were based on assumed T2 values: Asc, Asp, GABA, Gln, Gly, NAAG, PE, Lac + Thr, Glc + Tau.

Sixteen metabolites were quantified with the CRLBs below 30% in the occipital lobe, motor cortex, and cerebellum, and fifteen in basal ganglia, where it was not possible to quantify glycine. The CRLBs for the dominant metabolites (e.g. NAA, tCr, tCho, mIns, Glu) were found to be below 5% in all brain regions and were comparable to those obtained previously at 7 T in occipital lobe using either STEAM (11, 12) or SPECIAL (9). The CRLBs for other metabolites were below 30% and were comparable to those obtained previously in the occipital lobe (9, 11, 12) except for Gln, NAAG, and Asc which were higher in this study. The concentrations of metabolites obtained in the occipital lobe agreed well with previously published 7 T studies (9, 11, 12). The concentrations of two metabolites namely Gln and NAAG were lower in this study than in previously published work (9, 11, 12). For these metabolites, the T2 relaxation times were unknown in human brain at 7 T and were assumed to be similar to either Glu (in the case of Gln) or sNAA (in the case of NAAG).

Regional differences in the concentrations of the different metabolites were detected. The highest levels of tCr, tCho, Glc + Tau, Gln and mIns were detected in the cerebellum which agrees with previously published data (40, 41). The lowest level of Glu was detected in the motor cortex and the basal ganglia, while the highest level of GABA was detected in the basal ganglia. Metabolites such as GSH, sIns, and Lac + Thr were found to be uniformly distributed in all voxels.

Quantification of metabolites can be affected by T1 and T2 values of both water and metabolites. In the basal ganglia voxel, tissue heterogeneity is expected to lead to a range of water T1 values. Despite this, the effect of the uncertainty in T1 of water in the basal ganglia voxel on the measured concentration of metabolites was estimated to be only 2%. In the occipital cortex, lower than expected concentration of Gln was observed, which can be potentially caused by the fact that the same T2 relaxation time was used for both Glu and Gln since we were not able to measure the T2 of Gln in this study. Perhaps, there is a significant difference between T2 of those two molecules in the occipital cortex.

Metabolite quantification when using moderate to long echo time techniques necessitates measurement of relaxation times which has the disadvantage of increasing scan time, unless the relaxation times are known a priori and are known not to change with disease. In addition, in our approach the relaxation times of different moieties in the same molecule are assumed to be the same (e.g., Glu). However, T2 of metabolites has been shown to change in schizophrenia (42).

Conclusions

In this study, excellent quality spectra were obtained at 7 T using modified LASER sequence that enabled the quantification of fifteen metabolites in four regions of human brain. The quantification was performed taking into consideration CSF content, tissue composition, the relative density of MR-visible water, the T1 and T2 relaxation times of water in different tissue types, and the T2 relaxation of metabolites measured in this study. The observed concentrations in the occipital lobe agreed with previously reported values obtained using short echo time techniques. Regional differences in the T2 relaxation times as well as concentrations were detected and agreed with previous reports. The obtained results demonstrate that similar quality spectra can be obtained at 7 T in brain regions other than the occipital lobe and that quantification of longer echo time spectra is possible.

Acknowledgments

The authors thank Joanna Lukas and Stephen D. Weigand for help with statistics, Dinesh Deelchand, Pierre-Gilles Henry, Stéphane Lehéricy and Patrick Bolan for helpful discussion, and Jamie D. Walls for comments about the paper.

Sponsors: This work was supported by NIH: BTRC P41 RR008079, P30 NS057091, and the W.M. Keck Foundation.

List of abbreviations

AFP

adiabatic full passage

AHP

adiabatic half passage

Ala

alanine

ANOVA

analysis of variance

Asc

ascorbate

BG

basal ganglia

CR

cerebellum

Cr

creatine

CRLB

Cramér-Rao lower bounds

CSF

cerebrospinal fluid

FID

free induction decay

GABA

γ-aminobutyric acid

Glc

glucose

Gln

glutamine

Glu

glutamate

Gly

glycine

GM

gray matter

GPC

glycerophosphorylcholine

GSH

glutathione

Lac

lactate

LASER

localization by adiabatic selective refocusing

MC

motor cortex

mIns

myo-inositol

MM

macromolecules

mNAA

aspartyl resonances of NAA, NAA multiplet

NAA

N-acetylaspartate

NAAG

N-acetylaspartylglutamate

NS

number of scans

OC

occipital lobe

OVS

outer volume suppression

PCho

phosphorylcholine

PCr

phosphocreatine

PE

phosphorylethanolamine

RF

radio frequency

SD

standard deviation

sIns

scyllo-inositol

sNAA

methyl resonance of NAA, NAA singlet

SPECIAL

spin-echo, full intensity acquired localized

STEAM

stimulated echo acquisition mode

Tau

taurine

tCho

total choline, GPC + PCho

tCr

total creatine, creatine + phosphocreatine

TE

echo time

Thr

threonine

TIR

inversion time

TR

repetition time

TSE

turbo spin echo

VAPOR

variable pulse power and optimized relaxation delays

VOI

volume of interest

WM

white matter

References

  • 1.Pan JW, Twieg DB, Hetherington HP. Quantitative spectroscopic imaging of the human brain. Magn Reson Med. 1998;40(3):363–369. doi: 10.1002/mrm.1910400305. [DOI] [PubMed] [Google Scholar]
  • 2.Posse S, Otazo R, Caprihan A, Bustillo J, Chen HJ, Henry PG, Marjanska M, Gasparovic C, Zuo C, Magnotta V, Mueller B, Mullins P, Renshaw P, Ugurbil K, Lim KO, Alger JR. Proton echo-planar spectroscopic imaging of J-coupled resonances in human brain at 3 and 4 Tesla. Magn Reson Med. 2007;58(2):236–244. doi: 10.1002/mrm.21287. [DOI] [PubMed] [Google Scholar]
  • 3.Balchandani P, Pauly J, Spielman D. Interleaved narrow-band PRESS sequence with adiabatic spatial-spectral refocusing pulses for 1H MRSI at 7T. Magn Reson Med. 2008;59(5):973–979. doi: 10.1002/mrm.21539. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Hetherington HP, Avdievich NI, Kuznetsov AM, Pan JW. RF Shimming for Spectroscopic Localization in the Human Brain at 7 T. Magn Reson Med. 2010;63(1):9–19. doi: 10.1002/mrm.22182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Schuster C, Dreher W, Geppert C, Leibfritz D. Fast 3D 1H spectroscopic imaging at 3 tesla using spectroscopic missing-pulse SSFP with 3D spatial preselection. Magn Reson Med. 2007;57(1):82–89. doi: 10.1002/mrm.21105. [DOI] [PubMed] [Google Scholar]
  • 6.Schuster C, Dreher W, Stadler J, Bernarding J, Leibfritz D. Fast Three-Dimensional 1H MR Spectroscopic Imaging at 7 Tesla Using “Spectroscopic Missing Pulse - SSFP”. Magn Reson Med. 2008;60(5):1243–1249. doi: 10.1002/mrm.21787. [DOI] [PubMed] [Google Scholar]
  • 7.Scheenen TWJ, Heerschap A, Klomp DWJ. Towards 1H-MRSI of the human brain at 7T with slice-selective adiabatic refocusing pulses. Magn Reson Mater Phys. 2008;21:95–101. doi: 10.1007/s10334-007-0094-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Tkac I, Oz G, Adriany G, Ugurbil K, Gruetter R. In Vivo 1H NMR Spectroscopy of the Human Brain at High Magnetic Fields: Metabolite Quantification at 4T vs. 7T. Magn Reson Med. 2009;62(4):868–879. doi: 10.1002/mrm.22086. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Mekle R, Mlynarik V, Gambarota G, Hergt M, Krueger G, Gruetter R. MR Spectroscopy of the Human Brain With Enhanced Singla Intensity at Ultrashort Echo Times on a Clinical Platform at 3T and 7T. Magn Reson Med. 2009;61(6):1279–1285. doi: 10.1002/mrm.21961. [DOI] [PubMed] [Google Scholar]
  • 10.Tkac I, Andersen P, Adriany G, Merkle H, Ugurbil K, Gruetter R. In vivo 1H NMR spectroscopy of the human brain at 7 T. Magn Reson Med. 2001;46(3):451–456. doi: 10.1002/mrm.1213. [DOI] [PubMed] [Google Scholar]
  • 11.Mangia S, Tkac I, Gruetter R, Van de Moortele PF, Giove F, Maraviglia B, Ugurbil K. Sensitivity of single-voxel 1H-MRS in investigating the metabolism of the activated human visual cortex at 7 T. Magn Reson Imaging. 2006;24(4):343–348. doi: 10.1016/j.mri.2005.12.023. [DOI] [PubMed] [Google Scholar]
  • 12.Terpstra M, Ugurbil K, Tkac I. Noninvasive quantification of human brain ascorbate concentration using 1H NMR spectroscopy at 7 T. NMR Biomed. 2010;23(3):227–232. doi: 10.1002/nbm.1423. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Garwood M, DelaBarre L. The return of the frequency sweep: designing adiabatic pulses for contemporary NMR. J Magn Reson. 2001;153(2):155–177. doi: 10.1006/jmre.2001.2340. [DOI] [PubMed] [Google Scholar]
  • 14.Pan JW, Venkatraman T, Vives K, Spencer DD. Quantitative glutamate spectroscopic imaging of the human hippoicampus. NMR Biomed. 2006;19(2):209–216. doi: 10.1002/nbm.1019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Scheenen TWJ, Klomp DWJ, Wijnen JP, Heerschap A. Short echo time 1H-MRSI of the human brain at 3T with minimal chemical shift displacement errors using adiabatic refocusing pulses. Magn Reson Med. 2008;59(1):1–6. doi: 10.1002/mrm.21302. [DOI] [PubMed] [Google Scholar]
  • 16.Michaeli S, Garwood M, Zhu XH, DelaBarre L, Andersen P, Adriany G, Merkle H, Ugurbil K, Chen W. Proton T2 relaxation study of water, N-acetylaspartate, and creatine in human brain using Hahn and Carr-Purcell spin echoes at 4T and 7T. Magn Reson Med. 2002;47(4):629–633. doi: 10.1002/mrm.10135. [DOI] [PubMed] [Google Scholar]
  • 17.Xin L, Gambarota G, Mlynarik V, Gruetter R. Proton T2 relaxation time of J-coupled cerebral metabolites in rat brain at 9. 4T. NMR Biomed. 2008;21(4):396–401. doi: 10.1002/nbm.1205. [DOI] [PubMed] [Google Scholar]
  • 18.Adriany G, De Moortele PFV, Ritter J, Moeller S, Auerbach EJ, Akgun C, Snyder CJ, Vaughan T, Ugurbill K. A geometrically adjustable 16-channel transmit/receive transmission line array for improved RF efficiency and parallel imaging performance at 7 Tesla. Magn Reson Med. 2008;59(3):590–597. doi: 10.1002/mrm.21488. [DOI] [PubMed] [Google Scholar]
  • 19.Yousry TA, Schmid UD, Alkadhi H, Schmidt D, Peraud A, Buettner A, Winkler P. Localization of the motor hand area to a knob on the precentral gyrus - A new landmark. Brain. 1997;120:141–157. doi: 10.1093/brain/120.1.141. [DOI] [PubMed] [Google Scholar]
  • 20.Metzger GJ, Snyder C, Akgun C, Vaughan T, Ugurbil K, Van de Moortele PF. Local B1+ shimming for prostate imaging with transceiver arrays at 7T based on subject-dependent transmit phase measurements. Magn Reson Med. 2008;59(2):396–409. doi: 10.1002/mrm.21476. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Baum J, Tycko R, Pines A. Broadband and adiabatic inversion of a two-level system by phase-modulated pulses. Phys Rev A. 1985;32(6):3435–3447. doi: 10.1103/physreva.32.3435. [DOI] [PubMed] [Google Scholar]
  • 22.Silver MS, Joseph RI, Hoult DI. Highly selective π/2 and π pulse generation. J Magn Reson. 1984;59(3):347–351. [Google Scholar]
  • 23.Gruetter R. Automatic, localized in vivo adjustment of all first- and second-order shim coils. Magn Reson Med. 1993;29(6):804–811. doi: 10.1002/mrm.1910290613. [DOI] [PubMed] [Google Scholar]
  • 24.Gruetter R, Tkac I. Field mapping without reference scan using asymmetric echo-planar techniques. Magn Reson Med. 2000;43(2):319–323. doi: 10.1002/(sici)1522-2594(200002)43:2<319::aid-mrm22>3.0.co;2-1. [DOI] [PubMed] [Google Scholar]
  • 25.Tkac I, Starcuk Z, Choi IY, Gruetter R. In vivo 1H NMR spectroscopy of rat brain at 1 ms echo time. Magn Reson Med. 1999;41(4):649–656. doi: 10.1002/(sici)1522-2594(199904)41:4<649::aid-mrm2>3.0.co;2-g. [DOI] [PubMed] [Google Scholar]
  • 26.Pfeuffer J, Tkac I, Provencher SW, Gruetter R. Toward an in vivo neurochemical profile: quantification of 18 metabolites in short-echo-time 1H NMR spectra of the rat brain. J Magn Reson. 1999;141(1):104–120. doi: 10.1006/jmre.1999.1895. [DOI] [PubMed] [Google Scholar]
  • 27.Provencher SW. Estimation of metabolite concentrations from localized in vivo proton NMR spectra. Magn Reson Med. 1993;30(6):672–679. doi: 10.1002/mrm.1910300604. [DOI] [PubMed] [Google Scholar]
  • 28.Provencher SW. Automatic quantitation of localized in vivo 1H spectra with LCModel. NMR Biomed. 2001;14(4):260–264. doi: 10.1002/nbm.698. [DOI] [PubMed] [Google Scholar]
  • 29.Henry PG, Marjanska M, Walls JD, Valette J, Gruetter R, Ugurbil K. Proton-observed carbon-edited NMR spectroscopy in strongly coupled second-order spin systems. Magn Reson Med. 2006;55(2):250–257. doi: 10.1002/mrm.20764. [DOI] [PubMed] [Google Scholar]
  • 30.Govindaraju V, Young K, Maudsley AA. Proton NMR chemical shifts and coupling constants for brain metabolites. NMR Biomed. 2000;13(3):129–153. doi: 10.1002/1099-1492(200005)13:3<129::aid-nbm619>3.0.co;2-v. [DOI] [PubMed] [Google Scholar]
  • 31.Kaiser LG, Marjanska M, Matson GB, Iltis I, Bush SD, Soher BJ, Mueller S, Young K. 1H MRS detection of glycine residue of reduced glutathione in vivo. J Magn Reson. 2010;202(2):259–266. doi: 10.1016/j.jmr.2009.11.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Pijnappel WWF, Vandenboogaart A, Debeer R, Vanormondt D. Svd-Based Quantification of Magnetic-Resonance Signals. J Magn Reson. 1992;97(1):122–134. [Google Scholar]
  • 33.Gasparovic C, Song T, Devier D, Bockholt HJ, Caprihan A, Mullins PG, Posse S, Jung RE, Morrison LA. Use of tissue water as a concentration reference for proton spectroscopic imaging. Magn Reson Med. 2006;55(6):1219–1226. doi: 10.1002/mrm.20901. [DOI] [PubMed] [Google Scholar]
  • 34.Ernst T, Kreis R, Ross BD. Absolute Quantitation of Water and Metabolites in the Human Brain. 1. Compartments and Water. J Magn Reson B. 1993;102(1):1–8. [Google Scholar]
  • 35.Rooney WD, Johnson G, Li X, Cohen ER, Kim SG, Ugurbil K, Springer CS. Magnetic field and tissue dependencies of human brain longitudinal 1H2O relaxation in vivo. Magn Reson Med. 2007;57(2):308–318. doi: 10.1002/mrm.21122. [DOI] [PubMed] [Google Scholar]
  • 36.Bartha R, Michaeli S, Merkle H, Adriany G, Andersen P, Chen W, Ugurbil K, Garwood M. In vivo 1H2O T2(dagger) measurement in the human occipital lobe at 4T and 7T by Carr-Purcell MRI: Detection of microscopic susceptibility contrast. Magn Reson Med. 2002;47(4):742–750. doi: 10.1002/mrm.10112. [DOI] [PubMed] [Google Scholar]
  • 37.Kreis R, Slotboom J, Hofmann L, Boesch C. Integrated data acquisition and processing to determine metabolite contents, relaxation times, and macromolecule baseline in single examinations of individual subjects. Magn Reson Med. 2005;54(4):761–768. doi: 10.1002/mrm.20673. [DOI] [PubMed] [Google Scholar]
  • 38.Soher BJ, Pattany PM, Matson GB, Maudsley AA. Observation of coupled 1H metabolite resonances at long TE. Magn Reson Med. 2005;53(6):1283–1287. doi: 10.1002/mrm.20491. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Traber F, Block W, Lamerichs R, Gieseke J, Schild HH. 1H metabolite relaxation times at 3. 0 tesla: Measurements of T1 and T2 values in normal brain and determination of regional differences in transverse relaxation. J Magn Reson Imaging. 2004;19(5):537–545. doi: 10.1002/jmri.20053. [DOI] [PubMed] [Google Scholar]
  • 40.Michaelis T, Merboldt KD, Bruhn H, Hanicke W, Frahm J. Absolute Concentrations of Metabolites in the Adult Human Brain in Vivo - Quantification of Localized Proton MR Spectra. Radiology. 1993;187(1):219–227. doi: 10.1148/radiology.187.1.8451417. [DOI] [PubMed] [Google Scholar]
  • 41.Pouwels PJW, Frahm J. Regional metabolite concentrations in human brain as determined by quantitative localized proton MRS. Magn Reson Med. 1998;39(1):53–60. doi: 10.1002/mrm.1910390110. [DOI] [PubMed] [Google Scholar]
  • 42.Tunc-Skarka N, Weber-Fahr W, Hoerst M, Meyer-Lindenberg A, Zink M, Ende G. MR spectroscopic evaluation of N-acetylaspartate’s T2 relaxation time and concentration corroborates white matter abnormalities in schizophrenia. Neuroimage. 2009;48(3):525–531. doi: 10.1016/j.neuroimage.2009.06.061. [DOI] [PubMed] [Google Scholar]

RESOURCES