2D 1H spectroscopic imaging of the human brain at 4.1 T

Water and lipid suppression techniques for advanced 1 H MRS and MRSI of the human brain: Experts' consensus recommendations

The neurochemical information provided by proton magnetic resonance spectroscopy (MRS) or MR spectroscopic imaging (MRSI) can be severely compromised if strong signals originating from brain water and extracranial lipids are not properly suppressed. The authors of this paper present an overview of advanced water/lipid-suppression techniques and describe their advantages and disadvantages. Moreover, they provide recommendations for choosing the most appropriate techniques for proper use. Methods of water signal handling are primarily focused on the VAPOR technique and on MRS without water suppression (metabolite cycling). The section on lipid-suppression methods in MRSI is divided into three parts. First, lipid-suppression techniques that can be implemented on most clinical MR scanners (volume preselection, outer-volume suppression, selective lipid suppression) are described. Second, lipid-suppression techniques utilizing the combination of k-space filtering, high spatial resolutions and lipid regularization are presented. Finally, three promising new lipid-suppression techniques, which require special hardware (a multi-channel transmit system for dynamic B₁⁺ shimming, a dedicated second-order gradient system or an outer volume crusher coil) are introduced.

Fast 3D rosette spectroscopic imaging of neocortical abnormalities at 3 T: Assessment of spectral quality

PURPOSE

To use a fast 3D rosette spectroscopic imaging acquisition to quantitatively evaluate how spectral quality influences detection of the endogenous variation of gray and white matter metabolite differences in controls, and demonstrate how rosette spectroscopic imaging can detect metabolic dysfunction in patients with neocortical abnormalities.

METHODS

Data were acquired on a 3T MR scanner and 32-channel head coil, with rosette spectroscopic imaging covering a 4-cm slab of fronto-parietal-temporal lobes. The influence of acquisition parameters and filtering on spectral quality and sensitivity to tissue composition was assessed by LCModel analysis, the Cramer-Rao lower bound, and the standard errors from regression analyses. The optimized protocol was used to generate normative white and gray matter regressions and evaluate three patients with neocortical abnormalities.

RESULTS

As a measure of the sensitivity to detect abnormalities, the standard errors of regression for Cr/NAA and Ch/NAA were significantly correlated with the Cramer-Rao lower bound values (R = 0.89 and 0.92, respectively, both with P < 0.001). The rosette acquisition with a duration of 9.6 min, produces a mean Cramer-Rao lower bound (%) over the entire slab of 4.6 ± 2.6 and 5.8 ± 2.3 for NAA and Cr, respectively. This enables a Cr/NAA standard error of 0.08 (i.e., detection sensitivity of 25% for a 50/50 mixed gray and white matter voxel). In healthy controls, the regression of Cr/NAA versus fraction gray matter in the cingulate differs from frontal and parietal regions.

CONCLUSIONS

Fast rosette spectroscopic imaging acquisitions with regression analyses are able to identify metabolic differences across 4-cm slabs of the brain centrally and over the cortical periphery with high efficiency, generating results that are consistent with clinical findings. Magn Reson Med 79:2470-2480, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Improved spatial coverage for brain 3D PRESS MRSI by automatic placement of outer-volume suppression saturation bands

PURPOSE

To develop a technique for optimizing coverage of brain 3D (1) H magnetic resonance spectroscopic imaging (MRSI) by automatic placement of outer-volume suppression (OVS) saturation bands (sat bands) and to compare the performance for point-resolved spectroscopic sequence (PRESS) MRSI protocols with manual and automatic placement of sat bands.

MATERIALS AND METHODS

The automated OVS procedure includes the acquisition of anatomic images from the head, obtaining brain and lipid tissue maps, calculating optimal sat band placement, and then using those optimized parameters during the MRSI acquisition. The data were analyzed to quantify brain coverage volume and data quality.

RESULTS

3D PRESS MRSI data were acquired from three healthy volunteers and 29 patients using protocols that included either manual or automatic sat band placement. On average, the automatic sat band placement allowed the acquisition of PRESS MRSI data from 2.7 times larger brain volumes than the conventional method while maintaining data quality.

CONCLUSION

The technique developed helps solve two of the most significant problems with brain PRESS MRSI acquisitions: limited brain coverage and difficulty in prescription. This new method will facilitate routine clinical brain 3D MRSI exams and will be important for performing serial evaluation of response to therapy in patients with brain tumors and other neurological diseases.

Dual-band water and lipid suppression for MR spectroscopic imaging at 3 Tesla

A dual-band water and lipid suppression sequence was developed for multislice sensitivity-encoded proton MR spectroscopic imaging of the human brain. The presaturation scheme consisted of five dual-band frequency-modulated radiofrequency pulses based on hypergeometric functions integrated with eight outer volume suppression (OVS) pulses. The flip angles of the dual-band pulses were optimized through computer simulations to maximize suppression factors over a range of transmitter amplitude of radiofrequency field and water and lipid T(1) values. The resulting hypergeometric dual band with OVS (HGDB + OVS) sequence was implemented at 3 T in a multislice sensitivity-encoded proton MR spectroscopic imaging experiment and compared to a conventional water suppression scheme (variable pulse power and optimized relaxation delays (VAPOR)) with OVS. The HGDB sequence was significantly shorter than the VAPOR sequence (230 versus 728 msec). Both HGDB + OVS and VAPOR + OVS produced good water suppression, while lipid suppression with the HGDB + OVS sequence was far superior. In sensitivity-encoded proton MR spectroscopic imaging data, artifacts from extracranial lipid signals were significantly lower with HGDB + OVS. The shorter duration of HGDB compared to VAPOR also allows reduced pulse repetition time values in the multislice acquisition.

Macrophage delivery of therapeutic nanozymes in a murine model of Parkinson's disease

BACKGROUND

Parkinson's disease is a common progressive neurodegenerative disorder associated with profound nigrostriatal degeneration. Regrettably, no therapies are currently available that can attenuate disease progression. To this end, we developed a cell-based nanoformulation delivery system using the antioxidant enzyme catalase to attenuate neuroinflammatory processes linked to neuronal death.

METHODS

Nanoformulated catalase was obtained by coupling catalase to a synthetic polyelectrolyte of opposite charge, leading to the formation of a polyion complex micelle. The nanozyme was loaded into bone marrow macrophages and its transport to the substantia nigra pars compacta was evaluated in 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-intoxicated mice.

RESULTS

Therapeutic efficacy of bone marrow macrophages loaded with nanozyme was confirmed by twofold reductions in microgliosis as measured by CD11b expression. A twofold increase in tyrosine hydroxylase-expressing dopaminergic neurons was detected in nanozyme-treated compared with untreated MPTP-intoxicated mice. Neuronal survival was confirmed by magnetic resonance spectroscopic imaging. Bone marrow macrophage-loaded catalase showed sustained release of the enzyme in plasma.

CONCLUSION

These data support the importance of macrophage-based nanozyme carriage for Parkinson's disease therapies.

RF shimming for spectroscopic localization in the human brain at 7 T

Spectroscopic imaging of the human head at short echo times (<or=15 ms) typically requires suppression of signals from extracerebral tissues. However, at 7 T, decreasing efficiency in B1+ generation (hertz/watt) and increasing spectral bandwidth result in dramatic increases in power deposition and increased chemical shift registration artifacts for conventional gradient-based in-plane localization. In this work, we describe a novel method using radiofrequency shimming and an eight-element transceiver array to generate a B1+ field distribution that excites a ring about the periphery of the head and leaves central brain regions largely unaffected. We have used this novel B1+ distribution to provide in-plane outer volume suppression (>98% suppression of extracerebral lipids) without the use of gradients. This novel B1+ distribution is used in conjunction with a double inversion recovery method to provide suppression of extracerebral resonances with T1s greater than 400 ms, while having negligible effect on metabolite ratios of cerebral resonances with T1s>1000 ms. Despite the use of two adiabatic pulses, the high efficiency of the ring distribution allows radiofrequency power deposition to be limited to 3-4 W for a pulse repetition time of 1.5 sec. The short echo time enabled the acquisition of images of the human brain, displaying glutamate, glutamine, macromolecules, and other major cerebral metabolites.

Short echo spectroscopic imaging of the human brain at 7T using transceiver arrays

Recent advances in magnet technology have enabled the construction of ultrahigh-field magnets (7T and higher) that can accommodate the human head and body. Despite the intrinsic advantages of performing spectroscopic imaging at 7T, increased signal-to-noise ratio (SNR), and spectral resolution, few studies have been reported to date. This limitation is largely due to increased power deposition and B(1) inhomogeneity. To overcome these limitations, we used an 8-channel transceiver array with a short TE (15 ms) spectroscopic imaging sequence. Utilizing phase and amplitude mapping and optimization schemes, the 8-element transceiver array provided both improved efficiency (17% less power for equivalent peak B(1)) and homogeneity (SD(B(1)) = +/-10% versus +/-22%) in comparison to a transverse electromagnetic (TEM) volume coil. To minimize the echo time to measure J-modulating compounds such as glutamate, we developed a short TE sequence utilizing a single-slice selective excitation pulse followed by a broadband semiselective refocusing pulse. Extracerebral lipid resonances were suppressed with an inversion recovery pulse and delay. The short TE sequence enabled visualization of a variety of resonances, including glutamate, in both a control subject and a patient with a Grade II oligodendroglioma.

Magnetic resonance spectroscopy: principles and applications in neurosurgery

Although MRI is a routine and invaluable tool in diagnosis and presurgical planning, the related technique of magnetic resonance spectroscopy (MRS) is not often applied. MRS shows the chemical content of brain tissue and can therefore increase the specificity of diagnosis considerably. It also is able to detect abnormality in some tissues which appear normal on conventional MRI, with potential to aid in both discriminating lesion borders and in predicting post-operative outcome. The physical principles of MRS and its technical limitations are discussed, with examples of applications in tumors, epilepsy and traumatic brain injury.

Whole-brain N-acetylaspartate MR spectroscopic quantification: performance comparison of metabolite versus lipid nulling

BACKGROUND AND PURPOSE

Despite the prominent peak of N-acetylaspartate (NAA) in proton MR spectroscopy ((1)H-MR spectroscopy) of the adult brain and its almost exclusive presence in neuronal cells, the total amount of NAA, regarded as their marker, is difficult to obtain due to signal contamination from the skull lipids. This article compares the performance of 2 methods that overcome this difficulty to yield the whole-brain NAA signal, important for the assessment of the total disease load in diffuse neurologic disorders.

MATERIALS AND METHODS

The heads of 12 healthy volunteers, 3 women and 9 men, 31.0 +/- 7.1 years of age, were scanned at 3T by using 2 nonlocalizing (1)H-MR spectroscopy sequences: One nulls the NAA (TI = 940 ms) every second acquisition by inversion-recovery to cancel the signals of the lipids (T1 << TI) in an add-subtract scheme. The other nulls the signal of the lipids (TI = 155 ms) directly after each acquisition, requiring half as many averages for the same signal-to-noise ratio. Each sequence was repeated 3 times back-to-back on 3 occasions, and the comparison criteria were intrasubject precision (reproducibility) and total measurement duration.

RESULTS

NAA nulling is nearly twice as precise in its intrinsic back-to-back (5.8% versus 8.6%) as well as longitudinal (10.6% versus 19.7%) coefficients of variation compared with lipid nulling, but at the cost of double the acquisition time.

CONCLUSION

When speed is a more stringent requirement than precision, the new lipid-nulling sequence is a viable alternative. For precision in cross-sectional or longitudinal global NAA quantification, however, NAA nulling is still the approach of choice despite its x2 ( approximately 5 minutes) time penalty compared with the lipid-nulling approach.

Spectroscopic imaging of the pilocarpine model of human epilepsy suggests that early NAA reduction predicts epilepsy

Reduced hippocampal N-acetyl aspartate (NAA) is commonly observed in patients with advanced, chronic temporal lobe epilepsy (TLE). It is unclear, however, whether an NAA deficit is also present during the clinically quiescent latent period that characterizes early TLE. This question has important implications for the use of MR spectroscopic imaging (MRSI) in the early identification of patients at risk for TLE. To determine whether NAA is diminished during the latent period, we obtained high-resolution (1)H spectroscopic imaging during the latent period of the rat pilocarpine model of human TLE. We used actively detuneable surface reception and volume transmission coils to enhance sensitivity and a semiautomated voxel shifting method to accurately position voxels within the hippocampi. During the latent period, 2 and 7 d following pilocarpine treatment, hippocampal NAA was significantly reduced by 27.5 +/- 6.9% (P < 0.001) and 17.3 +/- 6.9% (P < 0.001) at 2 and 7 d, respectively. Quantitative estimates of neuronal loss at 7 d (2.3 +/- 7.7% reduction; P = 0.58, not significant) demonstrate that the NAA deficit is not due to neuron loss and therefore likely represents metabolic impairment of hippocampal neurons during the latent phase. Therefore, spectroscopic imaging provides an early marker for metabolic dysfunction in this model of TLE.

Quantitative 1H magnetic resonance spectroscopic imaging determines therapeutic immunization efficacy in an animal model of Parkinson's disease

Nigrostriatal degeneration, the pathological hallmark of Parkinson's disease (PD), is mirrored by 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) intoxication. MPTP-treated animals show the common behavioral, motor, and pathological features of human disease. We demonstrated previously that adoptive transfer of Copaxone (Cop-1) immune cells protected the nigrostriatal dopaminergic pathway in MPTP-intoxicated mice. Herein, we evaluated this protection by quantitative proton magnetic resonance spectroscopic imaging (1H MRSI). 1H MRSI performed in MPTP-treated mice demonstrated that N-acetyl aspartate (NAA) was significantly diminished in the substantia nigra pars compacta (SNpc) and striatum, regions most affected in human disease. When the same regions were coregistered with immunohistochemical stains for tyrosine hydroxylase, numbers of neuronal bodies and termini were similarly diminished. MPTP-intoxicated animals that received Cop-1 immune cells showed NAA levels, in the SNpc and striatum, nearly equivalent to PBS-treated animals. Moreover, adoptive transfer of immune cells from ovalbumin-immunized to MPTP-treated mice failed to alter NAA levels or protect dopaminergic neurons and their projections. These results demonstrate that 1H MRSI can evaluate dopaminergic degeneration and its protection by Cop-1 immunization strategies. Most importantly, the results provide a monitoring system to assess therapeutic outcomes for PD.

Coregistration of quantitative proton magnetic resonance spectroscopic imaging with neuropathological and neurophysiological analyses defines the extent of neuronal impairments in murine human immunodeficiency virus type-1 encephalitis

Relatively few immune-activated and virus-infected mononuclear phagocytes (MP; perivascular macrophages and microglia) may affect widespread neuronal dysfunction during human immunodeficiency virus type 1 (HIV-1)-associated dementia (HAD). Indeed, histopathological evidence of neuronal dropout often belies the extent of cognitive impairment. To define relationships between neuronal function and histopathology, proton magnetic resonance spectroscopic imaging (1H MRSI) and hippocampal long-term potentiation (LTP) were compared with neuronal and glial immunohistology in a murine model of HIV-1 encephalitis (HIVE). HIV-1(ADA)-infected human monocyte-derived macrophages (MDM) were stereotactically injected into the subcortex of severe combined immunodeficient (SCID) mice. Sham-operated and unmanipulated mice served as controls. Seven days after cell injection, brain histological analyses revealed a focal giant cell encephalitis, with reactive astrocytes, microgliosis, and neuronal dropout. Strikingly, significant reductions in N-acetyl aspartate concentration ([NAA]) and LTP levels in HIVE mice were in both injected and contralateral hemispheres and in brain subregions, including the hippocampus, where neuropathology was limited or absent. The data support the importance of 1H MRSI as a tool for assessing neuronal function for HAD. The data also demonstrate that a highly focal encephalitis can produce global deficits for neuronal function and metabolism.

Truncation artifact reduction in spectroscopic imaging using a dual-density spiral k-space trajectory

Truncation artifacts arise in magnetic resonance spectroscopic imaging (MRSI) of the human brain due to limited coverage of k-space necessitated by low SNR of metabolite signal and limited scanning time. In proton MRSI of the head, intense extra-cranial lipid signals "bleed" into brain regions, thereby contaminating signals of metabolites therein. This work presents a data acquisition strategy for reducing truncation artifact based on extended k-space coverage achieved with a dual-SNR strategy. Using the fact that the SNR in k-space increases monotonically with sampling density, dual-SNR is achieved in an efficient manner with a dual-density spiral k-space trajectory that permits a smooth transition from high density to low density. The technique is demonstrated to be effective in reducing "bleeding" of extra-cranial lipid signals while preserving the SNR of metabolites in the brain.

Regional cerebral blood flow and magnetic resonance spectroscopic imaging findings in diaschisis from stroke

BACKGROUND AND PURPOSE

This study evaluated blood flow and metabolite changes in cerebral diaschisis from internal capsule region infarction using regional cerebral blood flow (rCBF) single-photon emission computed tomography (SPECT) and 1H magnetic resonance spectroscopic imaging (MRSI). We hypothesized that complementary measures of diaschisis effects in white matter (characterized by 1H MRSI) and gray matter (characterized by changes in rCBF) can be measured and exhibit parallel changes.

METHODS

Five stroke patients and 16 normal controls underwent Tc-99m hexamethylpropyleneamine-oxime brain SPECT and 1H MRSI at 4.1 T. The metabolites N-acetyl aspartate (NAA) and creatine (Cr) were measured using 1H MRSI. The tissue content was expressed as the percent of gray or white matter in each MRSI voxel to allow comparison of the differential effects of diaschisis in gray and white matter tissue types. The blood flow and metabolite changes were evaluated at superior cerebral regions distant from the stroke to allow a measure of diaschisis relatively unconfounded by their expected changes in the infarction region.

RESULTS

The rCBF SPECT data in stroke patients showed a perfusion defect, with size ranging from 1.23 cc to 10.23 cc, in the region of cortical diaschisis. 1H MRSI showed increased Cr/NAA ratios in regions of white matter diaschisis. There was a tendency for larger rCBF defect size to be associated with greater increases in Cr/NAA values in the same diaschitic cerebral hemisphere, ipsilateral to the infarction.

CONCLUSION

Diaschisis ipsilateral to stroke in white matter can be characterized by 1H MRSI, and diaschisis ipsilateral to stroke in cortical gray matter regions can be characterized by changes in rCBF. The tendency for greater reductions in cortical rCBF values to be associated with increased Cr/NAA values in the same diaschitic cerebral hemisphere implies that a relationship exists between rCBF reductions in gray matter and abnormal changes in white matter subservient to it.

In vivo magnetic resonance spectroscopy and its application to neuropsychiatric disorders

In vivo magnetic resonance spectroscopy (MRS) is the only noninvasive imaging technique that can directly assess the living biochemistry in localized brain regions. In the past decade, spectroscopy studies have shown biochemical alterations in various neuropsychiatric disorders. These first-generation studies have, in most cases, been exploratory but have provided insightful biochemical information that has furthered our understanding of different brain disorders. This review provides a brief description of spectroscopy, followed by a literature review of key spectroscopy findings in schizophrenia, affective disorders, and autism. In schizophrenia, phosphorus spectroscopy studies have shown altered metabolism of membrane phospholipids (MPL) during the early course of the illness, which is consistent with a neurodevelopmental abnormality around the critical period of adolescence when the illness typically begins. Children and adolescents who are at increased genetic risk for schizophrenia show similar MPL alterations, suggesting that schizophrenia subjects with a genetic predisposition may have a premorbid neurodevelopmental abnormality. Independent of medication status, bipolar subjects in the depressive state tended to have higher MPL precursor levels and a deficit of high-energy phosphate metabolites, which also is consistent with major depression, though these results varied. Further bipolar studies are needed to investigate alterations at the early stage. Lastly, associations between prefrontal metabolism of high-energy phosphate and MPL and neuropsychological performance and reduced N-acetylaspartate in the temporal and cerebellum regions have been reported in individuals with autism. These findings are consistent with developmental alterations in the temporal lobe and in the cerebellum of persons with autism. This paper discusses recent findings of new functions of N-acetylaspartate.

Short echo time multislice proton magnetic resonance spectroscopic imaging in human brain: metabolite distributions and reliability

Multislice proton magnetic resonance spectroscopic imaging (1H MRSI) at 25 ms echo time was used to measure concentrations of myo-inositol (mI), N-acetylaspartate (NAA), and creatine (Cr) and choline (Cho) in ten normal subjects between 22 and 84 years of age (mean age 44 +/- 18 years). By co-analysis with MRI based tissue segmentation results, metabolite distributions were analyzed for each tissue type and for different brain regions. Measurement reliability was evaluated using intraclass correlation coefficients (ICC). Significant differences in metabolite distributions were found for all metabolites. mI of frontal gray matter was 84% of parietal gray matter and 87% of white matter. NAA of frontal gray matter was 86% of parietal gray matter and 85% of white matter. Cho of frontal gray matter was 125% of parietal gray matter and 59% of white matter and Cho of parietal gray matter was 47% of white matter. Cr of parietal gray matter was 113% of white matter. Reliability was relatively high (ICC from.70 to.93) for all metabolites in white matter and for NAA and Cr in gray matter, though limited (ICC less than.63) for mI and Cho in gray matter. These findings indicate that voxel gray/white matter contributions, regional variations in metabolite concentrations, and reliability limitations must be considered when interpreting 1H MR spectra of the brain.

Delayed-focus pulses optimized using simulated annealing

Unlike prefocused pulses and shaped pulses based on the linear response theory, delayed-focus pulses (X.-L. Wu et al., 1991, Magn. Reson. Med. 20, 165--170) produce a selective spin echo after a predefined short delay without using a pi refocusing pulse. In this paper, a series of delayed-focus pulses of different flip angles are proposed based on optimization using Fourier series representation and simulated annealing. The resistance of these delayed-focus pulses to T(2) relaxation is also demonstrated using numerical simulation of Bloch equations.

Short TE in vivo (1)H MR spectroscopic imaging at 1.5 T: acquisition and automated spectral analysis

Spectral analysis of short TE in vivo proton magnetic resonance spectroscopic imaging (MRSI) data are complicated by the presence of spectral overlap, low signal to noise and uncharacterized signal contributions. In this study, it is shown that an automated data analysis method can be used to generate metabolite images from MRSI data obtained from human brain at TE = 25 ms and 1.5 T when optimized pulse sequences and a priori metabolite knowledge are used. The analysis approach made use of computer simulation methods to obtain a priori spectral information of the metabolites of interest and utilized a combination of parametric spectral modeling and non-parametric signal characterization for baseline fitting. This approach was applied to data from optimized PRESS-SI and multi-slice spin-echo SI acquisitions, for which sample spectra and metabolite images are shown.

Nonlinear Trends in Hippocampal Metabolic Function and Verbal Memory: Evidence of Cognitive Reserve in Temporal Lobe Epilepsy?

The present study explored the possibility of nonlinear trends in the relationship between verbal memory and hippocampal function in a series of 33 patients with temporal lobe epilepsy (TLE). Right and left hippocampal metabolic function was quantified using levels of hippocampal creatine to N-acetylaspartate (Cr/NAA) derived from (1)H magnetic resonance spectroscopic imaging. An exploratory neural network analysis (multi-layer perceptron) suggested the possibility of either a quadratic or cubic trend in the relationship between left hippocampal Cr/NAA and verbal retention. Using regression-based curve estimation, the cubic function was found to optimally fit the data, explaining 41% of the variance in the relationship between verbal memory and hippocampal metabolic function. This was contrasted to the 28% variance explained by simple linear regression. These findings suggest that (1) the relationship between verbal retention and hippocampal function in patients with TLE is nonlinear, and (2) this could be explained in terms of a possible "cognitive reserve." Clinical implications are discussed.

Statistically driven identification of focal metabolic abnormalities in temporal lobe epilepsy with corrections for tissue heterogeneity using 1H spectroscopic imaging

1H spectroscopic imaging of N-acetyl-aspartate, creatine, and choline has proven to be a sensitive indicator for the lateralization of seizure foci in temporal lobe epilepsy. Previous studies have used right-left comparisons to identify the epileptogenic tissue assuming that alterations due to the disease process outweigh the effects of tissue heterogeneity. To evaluate the effectiveness of tissue heterogeneity corrected analyses, we evaluated three criteria for lateralization of the seizure focus: 1) a statistically driven method adjusted for tissue composition, 2) a single valued threshold, and 3) a single global index of the hippocampus. The statistically driven analysis lateralized all eight patients correctly, whereas the single threshold method incorrectly lateralized one case and the global index failed to identify a significant difference in two cases. These findings indicate that increased accuracy and sensitivity can be obtained by correcting for tissue heterogeneity when analyzing spectroscopy studies of temporal lobe epilepsy.

Spectroscopic imaging of the uptake kinetics of human brain ethanol

Previous measurements of the ratio of brain to venous blood alcohol have ranged from 21-100%, depending on the experimental model, pulse sequence, and the concentration reference used. The goal of this study was to evaluate the uptake kinetics and visibility of brain ethanol in comparison to venous blood levels using a pulse sequence that minimizes uncertainties due to differences in J-modulation, T(1), and T(2) between ethanol and the concentration standard. This was achieved using a short TE (24 msec) spin echo sequence with a semiselective refocusing pulse to minimize J-modulation losses of the ethanol. Brain ethanol levels were measured with 10-min time resolution using a 16 x 16 spectroscopic imaging matrix with nominal voxels of 1.44 cc. During the course of the study, the brain/blood alcohol ratio declined from a value of 1.54 +/- 0.74 at 35 min after drinking to a final value of 0.93 +/- 0.16 at 85 min postdrinking. Magn Reson Med 42:1019-1026, 1999.

Linear projection method for automatic slice shimming

A fast, reliable automatic slice shimming method is described. In-slice shim adjustments are based on one-dimensional phase mapping of four in-slice linear projections through the slice center. For axial, coronal, and sagittal slices it is shown that all in-slice first-, second-, and third-order spherical harmonic terms of B(0) inhomogeneity can be unequivocally determined and corrected. Through-slice shim adjustment is achieved using a one-dimensional projection of the entire slice or ROI along the slice-selection direction. Applications of this method to single-slice in vivo spectroscopic imaging of human brain have resulted in reproducible, high-quality spectroscopic data. Magn Reson Med 42:1082-1088, 1999.

Predictive value of 1H MRSI for outcome in temporal lobectomy

OBJECTIVE

To investigate the predictive value of 1H MRSI for outcome in patients with mesial temporal lobe epilepsy (MTLE).

BACKGROUND

1H MRSI has been shown to be highly sensitive in the lateralization of temporal lob epilepsy.

METHODS

The authors analyzed the relationship between the 1H MRSI findings and surgical outcome in 40 consecutive patients who underwent temporal lobe surgery for MTLE. Outcome at a mean of 24 months (range 18 to 40 months) was classified as seizure free or not seizure free.

RESULTS

At follow-up, 78% of patients were seizure free. Correlations showed no predictive value for the creatine/N-acetylated compound (Cr/NA) ratio of the operated temporal lobe and outcome. However, a relationship was found between surgical failure and the Cr/NA ratio of the nonoperated temporal lobe and with a Cr/NA ratio in the nonoperated lobe above 1.21 in patients with bilateral abnormalities (p < 0.01).

CONCLUSIONS

Preoperative elevations in the Cr/NA ratio in the nonoperated temporal lobe or the presence of higher metabolic ratios contralateral to the proposed surgery are associated with surgical failure. The predictive value of 1H MRSI absolute metabolite concentrations for outcome in MTLE requires further investigation.

A hybrid technique for spectroscopic imaging with reduced truncation artifact

Traditionally, Fourier spectroscopic imaging is associated with a small k-space coverage which leads to truncation artifacts such as "bleeding" and ringing in the resultant image. Because substantial truncation artifacts mainly arise from regions having intense signals, such as the subcutaneous lipid in the head, effective reduction of truncation artifacts can be achieved by obtaining an extended k-space coverage for these regions. In this paper, a hybrid technique which employs phase-encoded spectroscopic imaging (SI) to cover the central portion of the k-space and echo-planar spectroscopic imaging (EPSI) to measure the peripheral portion of the k-space is developed. EPSI, despite its inherently low SNR characteristics, provides a sufficient SNR for outer high-spatial frequency components of the aforementioned high signal regions and supplies an extended k-space coverage of these regions for the reduction of truncation artifacts. The data processing includes steps designed to remove inconsistency between the two types of data and a previously described technique for selectively retaining only outer k-space information for the high signal regions during the reconstruction. Experimental studies, in both phantoms and normal volunteers, demonstrate that the hybrid technique provides significant reduction in truncation artifacts.

Quantitative spectroscopic imaging of the human brain

A method to provide B1 correction and cerebrospinal fluid (CSF) referencing is developed and applied to spectroscopic imaging of the human brain at 4.1 T using a volume head coil. The B1 image allows rapid determination of the spatially dependent B1 that is then used to compensate for the B1 sensitivity of the spectroscopic sequence. The reference signal is acquired from CSF located in a lateral ventricular position using a point-resolved echo spectroscopy (PRESS) acquisition. The CSF spectrum is also corrected for B1 dependence. Together with T2 and T1 corrections, this method is used to provide quantitative values of N-acetylaspartate (NAA), creatine (Cr), and choline (Ch). The metabolite concentrations obtained from a spectroscopic imaging slice through the ventricles in seven normal controls are in good agreement with previously published literature values. This method is applied in a patient with secondary progressive multiple sclerosis, showing separate areas of abnormalities in both NAA and Cr.

Biological and clinical MRS at ultra-high field

The advantages of performing spectroscopic studies at higher field strengths include increased SNR, improved spectral resolution for J-coupled resonances, and improvements in the selectivity of spectral editing schemes. By using pulse sequences that minimize the required echo time, refocus J-evolution, employ low peak B1 requiring pulses and take advantage of spectroscopic imaging methods, these advantages can also be utilized in clinical applications of spectroscopy at high field. In addition to the static measurements measurements of N-acetyl aspartate (NAA), creatine (CR) and choline (CH) which can be performed at 1.5 T, high resolution measurements of glutamate, glutamine, GABA and the incorporation of 13C labeled glucose into glutamate are possible with improved spatial and spectral resolution. These methods have been utilized in patients with seizure disorders and multiple sclerosis to identify, characterize and map the metabolic changes associated with these diseases and their treatment.

High spatial resolution and speed in MRSI

The in vivo applications of magnetic resonance spectroscopic imaging (MRSI) have expanded significantly over the past 10 years and have reached the point where clinical trials are underway for a number of different diseases. One of the limiting factors in the widespread use of this technology has been the lack of widely available tools for obtaining data which are localized to sufficiently small tissue volumes to make an impact upon diagnosis and treatment planning. This is especially difficult within the timeframe of a clinical MR examination, which requires that both anatomic and metabolic data are acquired and processed. Recent advances in the hardware and software associated with clinical scanners have provided the potential for improvements in the spatial and time resolution of imaging and spectral data. The two areas which hold the most promise in terms of MRSI data are the use of phased array coils and the implementation of echo planar k-space sampling techniques. These could have immediate impact for 1H MRSI and may prove valuable for future applications of 31P MRSI.

Improved water and lipid suppression for 3D PRESS CSI using RF band selective inversion with gradient dephasing (BASING)

A T1 insensitive solvent suppression technique-band selective inversion with gradient dephasing (BASING)-was developed to suppress water and lipids for 1H magnetic resonance spectroscopy (MRS). BASING, which consists of a frequency selective RF inversion pulse surrounded by spoiler gradient pulses of opposite signs, was used to dephase stopband resonances and minimally impact passband metabolites. Passband phase linearity was achieved with a dual BASING scheme. Using the Shinnar-Le Roux algorithm, a highpass filter was designed to suppress water and rephase the lactate methyl doublet independently of TE, and water/lipid bandstop filters were designed for the brain and prostate. Phantom and in vivo experimental 3D PRESS CSI data were acquired at 1.5 T to compare BASING with CHESS and STIR suppression. With BASING, the measured suppression factor was over 100 times higher than with CHESS or STIR causing baseline distortions to be removed. It was shown that BASING can be incorporated into a variety of sequences to offer improved suppression in the presence of B1 and T1 inhomogeneites.

Normalization of contralateral metabolic function following temporal lobectomy demonstrated by 1H magnetic resonance spectroscopic imaging

We studied 10 medically intractable temporal lobe epilepsy (TLE) patients prior to surgery using proton magnetic resonance spectroscopic imaging (MRSI) to localize seizure foci. We found significantly elevated creatine/N-acetylaspartate (Cr/NAA) unilaterally in 8 and bilaterally in 2 patients. Five patients have been studied again 1 year after surgery. In the 2 patients with bilateral temporal seizure onsets, MRSI showed normalization of Cr/NAA in the unoperated contralateral tissue following surgical elimination of seizures. This study suggests that metabolic recovery can occur in contralateral temporal areas following surgical treatment of partial epilepsy.

Quantitative 1H spectroscopic imaging of human brain at 4.1 T using image segmentation

Metabolic differences in the content of N-acetylaspartate (NAA), creatinine (CR), and choline (CH) in cerebral gray and white matter can complicate the interpretation of 1H spectroscopic images. To account for these variations, the gray- and white-matter content of each voxel must be known. To provide these data, a T1-based image segmentation scheme was implemented at 4.1 T. The tissue composition of each voxel was determined using the point-spread function of the spectroscopic imaging acquisition and the segmented anatomical image. Pure gray- and white-matter values for CR/NAA and CH/NAA, and the content of CR, CH, and NAA, were determined using a linear-regression analysis of 984 voxels acquired from 10 subjects using white-matter CR as an internal standard. This information was used to establish means and confidence intervals for CR/NAA and CH/NAA from a voxel of arbitrary tissue composition. Using a single-tailed t test, the extent and locations of the metabolic abnormalities (P < 0.05) in a patient with multiple sclerosis were identified.

Evaluation of multiple sclerosis by 1H spectroscopic imaging at 4.1 T

The authors report on high-field (4.1 T) magnetic resonance 1H spectroscopic imaging studies on eight patients with relapsing remitting multiple sclerosis (mean expanded disability status scale (EDSS) 1.0) and eight normal controls. Using T1-weighted imaging to determine lesion position, the authors found the ratios of choline/N-acetyl (NA) compounds and creatine/NA were increased significantly in the multiple sclerosis (MS) patients relative to controls in lesioned tissue, adjacent to lesion, far removed from lesions as well as in periventricular tissue. The gray matter creatine/NA was mildly increased (P < 0.01) in the MS patients, whereas the elevated gray-matter ratio of choline/NA was of borderline significance (P = 0.13). A more detailed comparison of white-matter and mean gray-matter metabolite values indicates that creatine is increased greatest in areas far from lesions. This is in contrast to choline, which was greatest in lesions, and NA, which was smallest in lesions. It is postulated that the creatine increase may reflect an astrocytic (gliotic) or oligodendrocytic remyelinating process. The increased choline most likely reflects varying levels of inflammation and membrane turnover, whereas the NA decrease is representative of axonal dysfunction or loss.

Short echo time proton magnetic resonance spectroscopic imaging of macromolecule and metabolite signal intensities in the human brain

A novel approach is presented for imaging macromolecule and metabolite signals in brain by proton magnetic resonance spectroscopic imaging. The method differentiates between metabolites and macromolecules by T1 weighting using an inversion pulse followed by a variable inversion recovery time before localization and spectroscopic imaging. In healthy subjects, the major macromolecule resonances at 2.05 and 0.9 ppm were mapped at a nominal spatial resolution of 1 x 1 x 1.5 cm3 and were demonstrated to be highly reproducible between subjects. In subacute stroke patients, a highly elevated macromolecule resonance at 1.3 ppm was mapped to infarcted brain regions, suggesting potential applications for studying pathological conditions.

Proton nuclear magnetic resonance spectroscopic imaging of human temporal lobe epilepsy at 4.1 T

We performed proton magnetic resonance spectroscopic imaging (MRSI) at high magnetic field (4.1 T) to study N-acetylaspartate, creatine, and choline levels in the brains of normal control subjects and patients with intractable temporal lobe epilepsy. We compared the results of MRSI to those of other presurgical techniques to determine the sensitivity of this method in the lateralization of the epileptic focus. The normal hippocampal creatine-N-acetylaspartate ratio was 0.71 +/- 0.14 with no differences between left and right. Using the mean control hippocampal creatine-N-acetylaspartate ratio plus 2 standard deviations to identify statistically significant changes, we found lateralizing metabolic abnormalities corresponding to the operated temporal lobe in all patients. Four patients (40%) had contralateral abnormalities, and 2 of them had bilateral independent seizure onset confirmed by intracranial electroencephalographic studies. Statistically significant increases in the choline-N-acetylaspartate ratio in comparison to healthy volunteers were observed in 8 of the 10 patients. With the creatine-N-acetylaspartate ratio, MRSI demonstrated a 100% sensitivity compared to magnetic resonance imaging, which identified pathology in 70% of the patients. These findings suggest that proton MRSI yields a distinctive metabolic profile in patients with temporal lobe epilepsy and is sensitive in detecting bilateral metabolic abnormalities in some patients. These preliminary findings suggest that MRSI is more sensitive than magnetic resonance imaging in the lateralization of epileptic foci in temporal lobe epilepsy.

Three-dimensional spectroscopic imaging with time-varying gradients

A spectroscopic imaging sequence with a time-varying readout gradient in the slice selection direction is used to image multiple contiguous slices. For a given voxel size, the imaging time and signal-to-noise ratio of the three-dimensional spectroscopic sequence are the same as for a single slice acquisition without the oscillating readout gradient. The data reconstruction employs a gridding algorithm in two dimensions to interpolate the nonuniformly sampled data onto a Cartesian grid, and a fast Fourier transform in four dimensions: three spatial dimensions and the spectral dimension. The method is demonstrated by in vivo imaging of NAA in human brain at 1.5 T with 10 slices of 16 x 16 pixels spectroscopic images acquired in a total scan time of 17 min.

High resolution neuroimaging at 4.1T

In this article we report on acquisition of high resolution 512 x 512 images at 4.1T using an inversion recovery gradient-echo sequence and a volume head coil developed for high field applications. The Ti values for cerebral white and grey matter were measured to be 834 and 1282 ms, respectively. The partial saturation inversion recovery sequence (Tir 800 ms and TR 2500 ms) provided excellent contrast-to-noise for white to grey matter. Consequently, the images consistently visualized the thalamic nuclear groups, hippocampal fine structure, as well as small draining vessels of the white matter.

Application of high field spectroscopic imaging in the evaluation of temporal lobe epilepsy

Previous spectroscopic imaging studies of temporal lobe epilepsy have used comparisons of metabolite content or ratios to lateralize the seizure focus. Although highly successful, these studies have shown significant variations within each of the groups of healthy subjects and patients. This variation may arise from the natural differences seen in metabolite concentration in gray and white matter, the complex anatomy seen about the hippocampus, and the large voxels typically employed at 1.5 T. Using a 4.1 T whole body system, we have acquired spectroscopic images with 0.5 cc nominal voxels (1 cc after filtering) to evaluate the regional variation in metabolite content of the hippocampus, temporal gray and white matter, midbrain, and cerebellar vermis. Using a threshold value of 0.90 for CR/NAA, a value 90% of all normal hippocampal voxels lay below, we have correctly identified the presence of epileptogenic tissue in patients with unilateral as well as bilateral seizures. By using comparisons to healthy values of the CR/NAA ratio, this method enables the visualization of bilateral disease and provides information on the extent of gray matter involvement.