The performance of volume selection sequences for in vivo NMR spectroscopy: implications for quantitative MRS

Terminology and concepts for the characterization of in vivo MR spectroscopy methods and MR spectra: Background and experts' consensus recommendations

With a 40-year history of use for in vivo studies, the terminology used to describe the methodology and results of magnetic resonance spectroscopy (MRS) has grown substantially and is not consistent in many aspects. Given the platform offered by this special issue on advanced MRS methodology, the authors decided to describe many of the implicated terms, to pinpoint differences in their meanings and to suggest specific uses or definitions. This work covers terms used to describe all aspects of MRS, starting from the description of the MR signal and its theoretical basis to acquisition methods, processing and to quantification procedures, as well as terms involved in describing results, for example, those used with regard to aspects of quality, reproducibility or indications of error. The descriptions of the meanings of such terms emerge from the descriptions of the basic concepts involved in MRS methods and examinations. This paper also includes specific suggestions for future use of terms where multiple conventions have emerged or coexisted in the past.

GABA editing with macromolecule suppression using an improved MEGA-SPECIAL sequence

PURPOSE

The most common γ-aminobutyric-acid (GABA) editing approach, MEGA-PRESS, uses J-editing to measure GABA distinct from larger overlapping metabolites, but suffers contamination from coedited macromolecules (MMs) comprising 40 to 60% of the observed signal. MEGA-SPECIAL is an alternative method with better MM suppression, but is not widely used primarily because of its relatively poor spatial localization. Our goal was to develop an improved MM-suppressed GABA editing sequence at 3 Tesla.

METHODS

We modified a single-voxel MEGA-SPECIAL sequence with an oscillating readout gradient for improved spatial localization, and used very selective 30-ms editing pulses for improved suppression of coedited MMs.

RESULTS

Simulation and in vivo experiments confirmed excellent MM suppression, insensitive to the range of B₀ frequency drifts typically encountered in vivo. Both intersubject and intrasubject studies showed that MMs, when suppressed by the improved MEGA-SPECIAL method, contributed approximately 40% to the corresponding MEGA-PRESS measurements. From the intersubject study, the coefficient of variation for GABA+/Cre (MEGA-PRESS) was 11.2% versus 7% for GABA/Cre (improved MEGA-SPECIAL), demonstrating significantly reduced variance (P = 0.005), likely coming from coedited MMs.

CONCLUSIONS

This improved MEGA-SPECIAL sequence provides unbiased GABA measurements with reduced variance as compared with conventional MEGA-PRESS. This approach is also relatively insensitive to the range of B₀ drifts typically observed in in vivo human studies. Magn Reson Med 79:41-47, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Application of 31P MR spectroscopy to the brain tumors

Objective

To evaluate the clinical feasibility and obtain useful parameters of ³¹P magnetic resonance spectroscopy (MRS) study for making the differential diagnosis of brain tumors.

Materials and Methods

Twenty-eight patients with brain tumorous lesions (22 cases of brain tumor and 6 cases of abscess) and 11 normal volunteers were included. The patients were classified into the astrocytoma group, lymphoma group, metastasis group and the abscess group. We obtained the intracellular pH and the metabolite ratios of phosphomonoesters/phosophodiesters (PME/PDE), PME/inorganic phosphate (Pi), PDE/Pi, PME/adenosine triphosphate (ATP), PDE/ATP, PME/phosphocreatine (PCr), PDE/PCr, PCr/ATP, PCr/Pi, and ATP/Pi, and evaluated the statistical significances.

Results

The brain tumors had a tendency of alkalization (pH = 7.28 ± 0.27, p = 0.090), especially the pH of the lymphoma was significantly increased (pH = 7.45 ± 0.32, p = 0.013). The brain tumor group showed increased PME/PDE ratio compared with that in the normal control group (p = 0.012). The ratios of PME/PDE, PDE/Pi, PME/PCr and PDE/PCr showed statistically significant differences between each brain lesion groups (p < 0.05). The astrocytoma showed an increased PME/PDE and PME/PCr ratio. The ratios of PDE/Pi, PME/PCr, and PDE/PCr in lymphoma group were lower than those in the control group and astrocytoma group. The metastasis group showed an increased PME/PDE ratio, compared with that in the normal control group.

Conclusion

We have obtained the clinically applicable ³¹P MRS, and the pH, PME/PDE, PDE/Pi, PME/PCr, and PDE/PCr ratios are helpful for differentiating among the different types of brain tumors.

In vivo 31P spectroscopy by fully adiabatic extended image selected in vivo spectroscopy: a comparison between 3 T and 7 T

An improved image selected in vivo spectroscopy (ISIS) sequence for localized (31)P magnetic resonance spectroscopy at 7 T was developed. To reduce errors in localization accuracy, adiabatic excitation, gradient offset independent adiabatic inversion pulses, and a special extended ISIS ordering scheme were used. The localization accuracy of extended ISIS was investigated in phantoms. The possible spectral quality and reproducibility in vivo was explored in a volunteer (brain, muscle, and liver). A comparison between 3 T and 7 T was performed in five volunteers. Adiabatic extended ISIS provided high spectral quality and accurate localization. The contamination in phantom experiments was only ∼5%, even if a pulse repetition time ∼ 1.2·T(1) was chosen to maximize the signal-to-noise ratio per unit time. High reproducibility was found in the calf muscle for 2.5 cm isotropic voxels at 7 T. When compared with 3 T, localized (31)P magnetic resonance spectroscopy in the human calf muscle at 7 T provided ∼3.2 times higher signal-to-noise ratio (as judged from phosphocreatine peak amplitude in frequency domain after matched filtering). At 7 T, extended ISIS allowed the performance of high-quality localized (31)P magnetic resonance spectroscopy in a short measurement time (∼3 to 4 min) and isotropic voxel sizes of ∼2.5 to 3 cm. With such short measurement times, localized (31)P magnetic resonance spectroscopy has the potential to be applied not only for clinical research but also for routine clinical practice.

N-Acetyl peak in MR spectra of intracranial metastatic mucinous adenocarcinomas

Absence of N-acetylaspartate (NAA) is one important diagnostic criterion of MR spectroscopy (MRS) that may suggest that an intracranial mass lesion is a metastasis. We report two cases of histopathology-confirmed intracranial metastatic mucinous adenocarcinoma, which predominantly showed a large metabolite peak at 2.0 ppm, mimicking an NAA peak of normal brain tissue. This finding could be of help in the interpretation of MRS in cases of intracranial enhancing mass lesions, metastases or gliomas.

Limits of a localized magnetic resonance spectroscopy assay for ex vivo myocardial triacylglycerol

Localized magnetic resonance spectroscopy (LMRS) promises a powerful non-invasive means to determine myocardial triacylglycerol (TAG) in a clinical setting. Here, the linearity, specificity, robustness, precision, and accuracy of an ex vivo mouse-heart LMRS TAG assay are assessed by quantifying the spatial, spectral, and relaxation-induced uncertainties. The protocol, which is based on localization by adiabatic selective refocusing (LASER) using frequency offset corrected inversion (FOCI) pulses, alternating gradient polarity, and simple post-processing, is shown to have good characteristics. The presented protocol has a benchmark, phantom-based, accuracy of 3%, and when applied to ex vivo mouse hearts the accuracy is 6%, making the LMRS assay comparable to the typical destructive bioanalytical assay.

BSLIM: spectral localization by imaging with explicit B0 field inhomogeneity compensation

Magnetic resonance spectroscopy imaging (MRSI) is an attractive tool for medical imaging. However, its practical use is often limited by the intrinsic low spatial resolution and long acquisition time. Spectral localization by imaging (SLIM) has been proposed as a non-Fourier reconstruction algorithm that incorporates spatial a priori information about spectroscopically uniform compartments. Unfortunately, the influence of the magnetic field inhomogeneity--in particular, the susceptibility effects at tissues' boundaries--undermines the validity of the compartmental model. Therefore, we propose BSLIM as an extension of SLIM with field inhomogeneity compensation. A B0-field inhomogeneity map, which can be acquired rapidly and at high resolution, is used by the new algorithm as additional a priori information. We show that the proposed method is distinct from the generalized SLIM (GSLIM) framework. Experimental results of a two-compartment phantom demonstrate the feasibility of the method and the importance of inhomogeneity compensation.

Spatial localization in nuclear magnetic resonance spectroscopy

The ability to select a discrete region within the body for signal acquisition is a fundamental requirement of in vivo NMR spectroscopy. Ideally, it should be possible to tailor the selected volume to coincide exactly with the lesion or tissue of interest, without loss of signal from within this volume or contamination with extraneous signals. Many techniques have been developed over the past 25 years employing a combination of RF coil properties, static magnetic field gradients and pulse sequence design in an attempt to meet these goals. This review presents a comprehensive survey of these techniques, their various advantages and disadvantages, and implications for clinical applications. Particular emphasis is placed on the reliability of the techniques in terms of signal loss, contamination and the effect of nuclear relaxation and J-coupling. The survey includes techniques based on RF coil and pulse design alone, those using static magnetic field gradients, and magnetic resonance spectroscopic imaging. Although there is an emphasis on techniques currently in widespread use (PRESS, STEAM, ISIS and MRSI), the review also includes earlier techniques, in order to provide historical context, and techniques that are promising for future use in clinical and biomedical applications.

Valeurs contrôles obtenues avec une séquence press 135 ms en neurospectroscopie monovoxel du proton

The clinical value of MR spectroscopy is now well established and this technique has been added to the current French classification of medical acts (CCAM). This paper presents a set of normal control values for 3 metabolite ratios obtained using a PRESS sequence with a TE of 135 ms at 1.5T: NAA/Cho, NAA/Cr and Cho/CR. Spectroscopy data acquisition were obtained from the following 12 anatomical regions: parieto-occipital white matter, centrum semiovale, frontal white matter, thalamus, basal ganglia, cerebellum (hemisphere, including dentate nucleus), brain stem (including pons, medulla and midbrain), anterior and posterior temporal lobe, parietal, occipital and pre-frontal cortices. The presented data allow radiologists equipped with a similar MR system to implement a clinical spectroscopy program without undergoing research protocols in order to obtain control values.

1H magnetic resonance spectroscopy of invasive cervical cancer: an in vivo study with ex vivo corroboration

The objective of this study was to establish in vivo (1)H-magnetic resonance (MR) spectroscopic appearances of cervical cancer using an endovaginal receiver coil and corroborate findings with magic angle spinning (MAS) MR spectroscopy of tissue samples. Fifty-three women (14 controls and 39 with cervical cancer) underwent endovaginal coil MR imaging at 1.5 T with T(1)- and T(2)-weighted scans sagittal and transverse to the cervix. Localized (1)H MR spectra (PRESS technique, TR 1600 ms, TE 135 ms) were accumulated in all controls and 29 cancer patients whose tumour filled > 50% of a single 3.4 cm(3) voxel. Peaks from triglyceride-CH(2) and -CH(3) were defined as present and in-phase (with the choline resonance), present but out-of-phase, or not present. Peak areas of choline-containing compounds were standardized to the area of unsuppressed tissue water resonance. Comparisons in observed resonances between groups were made using Fisher's exact test (qualitative data) and a t-test (quantitative data). Biopsies from these women analysed using MAS-MR spectroscopy and normalized to the intensity of an external standard of silicone rubber were similarly compared. Adequate water suppression permitted spectral analysis in 11 controls and 27 cancer patients. In-phase triglyceride-CH(2) resonances (1.3 ppm) were observed in 74% of tumours but in no control women (p < 0.001). No differences were observed in the presence of a 2 ppm resonance, choline-containing compounds or creatine in cancer compared with control women. However, ex vivo analysis showed significant differences not only in -CH(2), but also in -CH(3), a 2 ppm resonance, choline-containing compounds and creatine between tissues from control women and cancer tissue (p < 0.001, = 0.001, = 0.036, < 0.001 and = 0.004 respectively). On in vivo (1)H-MR spectroscopy, the presence of positive triglyceride-CH(2) resonances can be used to detect and confirm the presence of cervical cancer. However, technical improvements are required before routine clinical use.

1H magnetic resonance spectroscopy of preinvasive and invasive cervical cancer: In vivo-ex vivo profiles and effect of tumor load

PURPOSE

To compare in vivo (1)H magnetic resonance (MR) spectra of preinvasive and invasive cervical lesions with ex vivo magic angle spinning (MAS) spectra of intact biopsies from the same subjects and to establish the effects of tumor load in the tissue sampled on the findings.

MATERIALS AND METHODS

A total of 51 subjects (nine with normal cervix, 10 with cervical intraepithelial neoplasia [CIN], and 32 with cervical cancer) underwent endovaginal MR at 1.5 T. Single-voxel (3.4 cm(3)) (1)H MR spectra were acquired and voxel tumor load was calculated (tumor volume within voxel as a percentage of voxel volume). Resonances from triglycerides -CH(2) and -CH(3) and choline-containing compounds (Cho) were correlated with voxel tumor load. Biopsies analyzed by (1)H MAS-MR spectroscopy (MRS) had metabolite levels correlated with tumor load in the sample at histology.

RESULTS

In vivo studies detected Cho in normal, CIN, and cancer patients with no significant differences in levels (P = 0.93); levels were independent of voxel tumor load. Triglyceride -CH(2) and -CH(3) signals in-phase with Cho were present in 77% and 29%, respectively, of cancer subjects (but not in normal women or those with CIN), but did not correlate with voxel tumor load. Ex vivo cancer biopsies showed levels of triglycerides -CH(2) and -CH(3) and of Cho that were significantly greater than in normal or CIN biopsies (P < 0.05); levels were independent of the tumor load in the sample. The presence of -CH(2) in vivo predicted the presence of cancer with a sensitivity and specificity of 77.4% and 93.8% respectively, positive (PPV) and negative (NPV) predictive values were 96% and 68.2%; for -CH(2) ex vivo, sensitivity was 100%; specificity, 69%; PPV, 82%; and NPV, 100%.

CONCLUSION

Elevated lipid levels are detected by MRS in vivo and ex vivo in cervical cancer and are independent of tumor load in the volume of tissue sampled.

Relaxation times of 31P-metabolites in human calf muscle at 3 T

Localized (31)P-STEAM experiments were performed at 3 T to estimate relaxation times of phosphorus-containing metabolites in the human calf muscle in vivo. T(1) and T(2) times of PCr, P(i), and NTPs were measured in the resting calf muscle of healthy subjects by varying TR and TE. The localization performance of the (31)P-STEAM sequence was evaluated on a test object, resulting in a relative selection efficiency of 78 +/- 1% and contamination from outside the voxel of 0 +/- 2% under fully relaxed conditions. T(1) relaxation times (+/-SD, n = 5) of P(i), PCr, gamma-NTP, alpha-NTP, and beta-NTP obtained at 3 T are 5.2 +/- 1.0 s, 6.4 +/- 0.2 s, 4.5 +/- 0.3 s, 2.6 +/- 0.9 s, and 3.5 +/- 1.1 s, respectively. T(2) relaxation times (+/-SD, n = 6) of these metabolites are 148 +/- 17 ms, 334 +/- 30 ms, 78 +/- 13 ms, 55 +/- 7 ms, and 55 +/- 10 ms, respectively. Spin-lattice relaxation times established at 3 T are consistent with literature data at lower field strengths, whereas spin-spin relaxation times are lower. Several methodological considerations are discussed which may help improve quantification of metabolite concentrations in the human (calf) muscle in vivo by using localized noninvasive (31)P-MRS at 3 T, which is currently being tested for routine clinical applications.