Direct in vivo measurement of human cerebral GABA concentration using MEGA-editing at 7 Tesla

Advances in high-field magnetic resonance imaging

Among advances in magnetic resonance imaging (MRI), the increase of the magnetic field strength is perhaps one of the most significant. The use of high magnetic fields for in vivo magnetic resonance is motivated by a number of considerations. Advantages are increases in signal-to-noise ratio, blood-oxygenation level-dependent contrast, and spectral resolution, while disadvantages include potential reduction of contrast in anatomic imaging owing to lengthening of T1 and effects of susceptibility of high fields. To address these challenges, technical advances have been made in various aspects of MRI, allowing high-field MRI to provide exquisite morphological and functional details in clinical and research settings. This review provides an overview of technical issues and applications of high-field MRI.

A new strategy for in vivo spectral editing. Application to GABA editing using selective homonuclear polarization transfer spectroscopy

A novel single-shot in vivo spectral editing method is proposed in which the signal to be detected, is regenerated anew from the thermal equilibrium magnetization of a source to which it is J-coupled. The thermal equilibrium magnetization of the signal to be detected together with those of overlapping signals are suppressed by single-shot gradient dephasing prior to the signal regeneration process. Application of this new strategy to in vivo GABA editing using selective homonuclear polarization transfer allows complete suppression of overlapping creatine and glutathione while detecting the GABA-4 methylene resonance at 3.02 ppm with an editing yield similar to that of conventional editing methods. The NAA methyl group at 2.02 ppm was simultaneously detected and can be used as an internal navigator echo for correcting the zero order phase and frequency shifts and as an internal reference for concentration. This new method has been demonstrated for robust in vivo GABA editing in the rat brain and for study of GABA synthesis after acute vigabatrin administration.

High-field proton MRS of human brain

Proton magnetic resonance spectroscopy (1H-MRS) of the brain reveals specific biochemical information about cerebral metabolites, which may support clinical diagnoses and enhance the understanding of neurological disorders. The advantages of performing 1H-MRS at higher field strengths include better signal to noise ratio (SNR) and increased spectral, spatial and temporal resolution, allowing the acquisition of high quality, easily quantifiable spectra in acceptable imaging times. In addition to improved measurement precision of N-acetylaspartate, choline, creatine and myo-inositol, high-field systems allow the high-resolution measurement of other metabolites, such as glutamate, glutamine, gamma-aminobutyric acid, scyllo-inositol, aspartate, taurine, N-acetylaspartylglutamate, glucose and branched amino acids, thus extending the range of metabolic information. However, these advantages may be hampered by intrinsic field-dependent technical difficulties, such as decreased T2 signal, chemical shift dispersion errors, J-modulation anomalies, increased magnetic susceptibility, eddy current artifacts, limitations in the design of homogeneous and sensitive radiofrequency (RF) coils, magnetic field instability and safety issues. Several studies demonstrated that these limitations could be overcome, suggesting that the appropriate optimization of high-field 1H-MRS would expand the application in the fields of clinical research and diagnostic routine.

1H NMR detection of vitamin C in human brain in vivo

Vitamin C (ascorbate) is well established as an essential nutrient that functions as an antioxidant. Since it is present in the human brain at detectable concentrations, this study was designed to detect and quantify ascorbate in the human brain in vivo using 1H NMR spectroscopy (MRS). Ascorbate was consistently detected in all five study subjects, and was measured using MEGA-PRESS difference editing. The in vivo resonance pattern was consistent with that of ascorbate based on position, line width, peak pattern, and relative intensity. Metabolites with a potential for coediting were assessed using phantom solutions. The putative resonances of myo-inositol, lactate, glycerophosphocholine, phosphocholine, and phosphoethanolamine were detected at positions distinct from those of ascorbate. This study represents the first in vivo detection of vitamin C in the human brain using 1H MRS. A concentration of 1.3 +/- 0.3 micromol/g (mean +/- SD, N = 4) was estimated.

Single-shot two-echo technique for simultaneous measurement of GABA and creatine in the human brain in vivo

A single-shot, two-echo method for the simultaneous detection of multiple-quantum (MQ)-filtered gamma-aminobutyric acid (GABA) and creatine (Cr) was developed and demonstrated in the human brain in vivo at 3 Tesla. The simultaneously measured Cr singlet served as a navigator for the spectral phase of GABA and any frequency shift during measurements due to drift in the static magnetic field (B(0)) or subject movement, as well as an internal concentration reference. In addition, the use of a double-band frequency-selective MQ filter for C(3) and C(4) methylene protons of GABA provided a very robust measurement of GABA, with excellent suppression of overlapping metabolites such as Cr and glutathione (GSH) in each single scan. Contamination from overlapping macromolecules was also demonstrated to be negligible with this method. The GABA-to-Cr ratio was 0.09 +/- 0.03 (mean +/- SD, N = 17) and the estimated concentration of GABA in the frontoparietal region of the human brain in vivo was 0.66 +/- 0.19 micromol/g (mean +/- SD, N = 17) with the internal reference method, and 0.69 +/- 0.18 micromol/g (mean +/- SD, N = 17) with the external reference method. The observed pattern of GABA doublet was consistent among all subjects, with a frequency separation of approximately 13 Hz.

In vivo GABA detection with improved selectivity and sensitivity by localized double quantum filter technique at 4.1T

Gamma-aminobutyric acid (GABA) is the major inhibitory neurotransmitter for the normal function of mammal and human brain. It is difficult to detect GABA signal with the conventional single quantum technique due to its relatively low concentration and overlapping with other signals from creatine (Cr), glutathione (GSH), as well as macromolecules. Using a high-selective read pulse, DANTE, and at the facility of increased sensitivity and chemical shift resolution at high-field 4.1T, GABA editing by double quantum filter (DQF) with robust suppression of Cr and GSH was achieved. Our editing efficiency of 40-50% was achievable on a GABA phantom (50 mM GABA and 61 mM choline). Furthermore, GABA editing spectra were acquired with echo time TE = 77 ms, and any possible macromolecular contamination to GABA editing spectra was found to be negligible. This high-field DQF setup was applied to 11 healthy volunteers, and the mean GABA level was measured to be 1.12 +/- 0.15 mM in the occipital lobe in reference to 7.1 mM Cr concentration.

In vivo 1H NMR spectroscopy of individual human brain metabolites at moderate field strengths

This article reviews spectral editing techniques for in vivo 1H NMR spectroscopy of human brain tissue at moderate field strengths of 1.5-3 Tesla. Various aspects of 1H NMR spectroscopy are discussed with regard to in vivo applications. The parameter set [delta, J, n] (delta being the relative chemical shift, J the scalar coupling constant and n the number of coupled spins) is used to characterize the spin systems under investigation and to classify the editing techniques that are used in in vivo 1H NMR spectroscopy.

Measurement of reduced glutathione (GSH) in human brain using LCModel analysis of difference-edited spectra

The concentration of reduced glutathione (GSH), an antioxidant, may be altered in various brain diseases. MEGA-PRESS was used to edit for the (1)H NMR signal from GSH in the occipital lobe of 12 normal humans. In all studies, GSH was clearly detected with a spectral pattern consistent with spectra acquired from a phantom containing GSH. Retention of singlet resonances in the subspectra, a key advantage of this difference-editing technique, provided an unambiguous reference for the offset and phase of the edited signal. Linear combination model (LCModel) analysis provided an unbiased means for quantifying signal contribution from edited metabolites. GSH concentration was estimated from the in vivo spectra as 1.3 +/- 0.2 micro mol/g (mean +/- SD, n = 12).