Effects of noise and linewidth on in vivo analysis of glutamate at 3 T

Comparison of baseline correction algorithms for in vivo 1H-MRS

Proton MRS is used clinically to collect localized, quantitative metabolic data from living tissues. However, the presence of baselines in the spectra complicates accurate MRS data quantification. The occurrence of baselines is not specific to short-echo-time MRS data. In short-echo-time MRS, the baseline consists typically of a dominating macromolecular (MM) part, and can, depending on B0 shimming, poor voxel placement, and/or localization sequences, also contain broad water and lipid resonance components, indicated by broad components (BCs). In long-echo-time MRS, the MM part is usually much smaller, but BCs may still be present. The sum of MM and BCs is denoted by the baseline. Many algorithms have been proposed over the years to tackle these artefacts. A first approach is to identify the baseline itself in a preprocessing step, and a second approach is to model the baseline in the quantification of the MRS data themselves. This paper gives an overview of baseline handling algorithms and also proposes a new algorithm for baseline correction. A subset of suitable baseline removal algorithms were tested on in vivo MRSI data (semi-LASER at TE = 40 ms) and compared with the new algorithm. The baselines in all datasets were removed using the different methods and subsequently fitted using spectrIm-QMRS with a TDFDFit fitting model that contained only a metabolite basis set and lacked a baseline model. The same spectra were also fitted using a spectrIm-QMRS model that explicitly models the metabolites and the baseline of the spectrum. The quantification results of the latter quantification were regarded as ground truth. The fit quality number (FQN) was used to assess baseline removal effectiveness, and correlations between metabolite peak areas and ground truth models were also examined. The results show a competitive performance of our new proposed algorithm, underscoring its automatic approach and efficiency. Nevertheless, none of the tested baseline correction methods achieved FQNs as good as the ground truth model. All separately applied baseline correction methods introduce a bias in the observed metabolite peak areas. We conclude that all baseline correction methods tested, when applied as a separate preprocessing step, yield poorer FQNs and biased quantification results. While they may enhance visual display, they are not advisable for use before spectral fitting.

Linewidth-related bias in modelled concentration estimates from GABA-edited 1H-MRS

J-difference-edited MRS is widely used to study GABA in the human brain. Editing for low-concentration target molecules (such as GABA) typically exhibits lower signal-to-noise ratio (SNR) than conventional non-edited MRS, varying with acquisition region, volume and duration. Moreover, spectral lineshape may be influenced by age-, pathology-, or brain-region-specific effects of metabolite T2, or by task-related blood-oxygen level dependent (BOLD) changes in functional MRS contexts. Differences in both SNR and lineshape may have systematic effects on concentration estimates derived from spectral modelling. The present study characterises the impact of lineshape and SNR on GABA+ estimates from different modelling algorithms: FSL-MRS, Gannet, LCModel, Osprey, spant and Tarquin. Publicly available multi-site GABA-edited data (222 healthy subjects from 20 sites; conventional MEGA-PRESS editing; TE = 68 ms) were pre-processed with a standardised pipeline, then filtered to apply controlled levels of Lorentzian and Gaussian linebroadening and SNR reduction. Increased Lorentzian linewidth was associated with a 2-5% decrease in GABA+ estimates per Hz, observed consistently (albeit to varying degrees) across datasets and most algorithms. Weaker, often opposing effects were observed for Gaussian linebroadening. Variations are likely caused by differing baseline parametrization and lineshape constraints between models. Effects of linewidth on other metabolites (e.g., Glx and tCr) varied, suggesting that a linewidth confound may persist after scaling to an internal reference. These findings indicate a potentially significant confound for studies where linewidth may differ systematically between groups or experimental conditions, e.g. due to T2 differences between brain regions, age, or pathology, or varying T2* due to BOLD-related changes. We conclude that linewidth effects need to be rigorously considered during experimental design and data processing, for example by incorporating linewidth into statistical analysis of modelling outcomes or development of appropriate lineshape matching algorithms.

Improving magnetic resonance spectroscopy in the brainstem periaqueductal gray using spectral registration

PURPOSE

Functional understanding of the periaqueductal gray (PAG), a clinically relevant brainstem region, can be advanced using 1 H-MRS. However, the PAG's small size and high levels of physiological noise are methodologically challenging. This study aimed to (1) improve 1 H-MRS quality in the PAG using spectral registration for frequency and phase error correction; (2) investigate whether spectral registration is particularly useful in cases of greater head motion; and (3) examine metabolite quantification using literature-based or individual-based water relaxation times.

METHODS

Spectra were acquired in 33 healthy volunteers (50.1 years, SD = 17.19, 18 females) on a 3 T Philipps MR system using a point-resolved spectroscopy (PRESS) sequence optimized with very selective saturation pulses (OVERPRESS) and voxel-based flip angle calibration (effective volume of interest size: 8.8 × 10.2 × 12.2 mm3 ). Spectra were fitted using LCModel and SNR, NAA peak linewidths and Cramér-Rao lower bounds (CRLBs) were measured after spectral registration and after minimal frequency alignment.

RESULTS

Spectral registration improved SNR by 5% (p = 0.026, median value post-correction: 18.0) and spectral linewidth by 23% (p < 0.001, 4.3 Hz), and reduced the metabolites' CRLBs by 1% to 15% (p < 0.026). Correlational analyses revealed smaller SNR improvements with greater head motion (p = 0.010) recorded using a markerless motion tracking system. Higher metabolite concentrations were detected using individual-based compared to literature-based water relaxation times (p < 0.001).

CONCLUSION

This study demonstrates high-quality 1 H-MRS acquisition in the PAG using spectral registration. This shows promise for future 1 H-MRS studies in the PAG and possibly other clinically relevant brain regions with similar methodological challenges.

Denoising single MR spectra by deep learning: Miracle or mirage?

PURPOSE

The inherently poor SNR of MRS measurements presents a significant hurdle to its clinical application. Denoising by machine or deep learning (DL) was proposed as a remedy. It is investigated whether such denoising leads to lower estimate uncertainties or whether it essentially reduces noise in signal-free areas only.

METHODS

Noise removal based on supervised DL with U-nets was implemented using simulated 1 H MR spectra of human brain in two approaches: (1) via time-frequency domain spectrograms and (2) using 1D spectra as input. Quality of denoising was evaluated in three ways: (1) by an adapted fit quality score, (2) by traditional model fitting, and (3) by quantification via neural networks.

RESULTS

Visually appealing spectra were obtained; hinting that denoising is well-suited for MRS. However, an adapted denoising score showed that noise removal is inhomogeneous and more efficient for signal-free areas. This was confirmed by quantitative analysis of traditional fit results as well as DL quantitation following DL denoising. DL denoising, although apparently successful as judged by mean squared errors, led to substantially biased estimates in both implementations.

CONCLUSION

The implemented DL-based denoising techniques may be useful for display purposes, but do not help quantitative evaluations, confirming expectations based on estimation theory: Cramér Rao lower bounds defined by the original data and the appropriate fitting model cannot be circumvented in an unbiased way for single data sets, unless additional prior knowledge can be incurred in the form of parameter restrictions/relations or applicable substates.

In vivo magnetic resonance spectroscopy by transverse relaxation encoding with narrowband decoupling

Cell pathology in neuropsychiatric disorders has mainly been accessible by analyzing postmortem tissue samples. Although molecular transverse relaxation informs local cellular microenvironment via molecule-environment interactions, precise determination of the transverse relaxation times of molecules with scalar couplings (J), such as glutamate and glutamine, has been difficult using in vivo magnetic resonance spectroscopy (MRS) technologies, whose approach to measuring transverse relaxation has not changed for decades. We introduce an in vivo MRS technique that utilizes frequency-selective editing pulses to achieve homonuclear decoupled chemical shift encoding in each column of the acquired two-dimensional dataset, freeing up the entire row dimension for transverse relaxation encoding with J-refocusing. This results in increased spectral resolution, minimized background signals, and markedly broadened dynamic range for transverse relaxation encoding. The in vivo within-subject coefficients of variation for the transverse relaxation times of glutamate and glutamine, measured using the proposed method in the human brain at 7 T, were found to be approximately 4%. Since glutamate predominantly resides in glutamatergic neurons and glutamine in glia in the brain, this noninvasive technique provides a way to probe cellular pathophysiology in neuropsychiatric disorders for characterizing disease progression and monitoring treatment response in a cell type-specific manner in vivo.

Neurochemical correlations in short echo time proton magnetic resonance spectroscopy

Neurochemical concentrations determined by magnetic resonance spectroscopy (MRS) have been treated as statistically independent measurements in various clinical MRS studies. However, spectral overlap, independent of any biological effects, could lead to significant correlations between neurochemical concentrations extracted from spectral fitting of MRS data, confounding determination of correlations of biological origin. Short echo time (TE) proton MRS spectra are very crowded because of the comparatively narrow chemical shift dispersion of proton nuclear spins. In this study, the complex neurochemical correlations of spectral origin in short-TE MRS spectra were quantified. The effects of macromolecules and the background spectral baseline on metabolite-metabolite correlations were also analyzed. Our results demonstrate the importance of factoring in spectral correlations when correlating overlapping metabolite signals in short-TE spectra with clinical parameters.

Genetically encoded fluorescent sensors for imaging neuronal dynamics in vivo

The brain relies on many forms of dynamic activities in individual neurons, from synaptic transmission to electrical activity and intracellular signaling events. Monitoring these neuronal activities with high spatiotemporal resolution in the context of animal behavior is a necessary step to achieve a mechanistic understanding of brain function. With the rapid development and dissemination of highly optimized genetically encoded fluorescent sensors, a growing number of brain activities can now be visualized in vivo. To date, cellular calcium imaging, which has been largely used as a proxy for electrical activity, has become a mainstay in systems neuroscience. While challenges remain, voltage imaging of neural populations is now possible. In addition, it is becoming increasingly practical to image over half a dozen neurotransmitters, as well as certain intracellular signaling and metabolic activities. These new capabilities enable neuroscientists to test previously unattainable hypotheses and questions. This review summarizes recent progress in the development and delivery of genetically encoded fluorescent sensors, and highlights example applications in the context of in vivo imaging.

Magnetic Field Dependence of Spectral Correlations between 31P-Containing Metabolites in Brain

Spectral correlations between metabolites in ³¹P magnetic resonance spectroscopy (MRS) spectra of human brain were compared at 3 and 7 Tesla, the two commonly used magnetic field strengths for clinical research. It was found that at both field strengths, there are significant correlations between ³¹P-containing metabolites arising from spectral overlap, and their downfield correlations are markedly altered by the background spectral baseline. Overall, the spectral correlations between ³¹P-containing metabolites are markedly reduced at 7 Tesla with the increased chemical shift dispersion and the decreased membrane phospholipid signal. The findings provide the quantitative landscape of pre-existing correlations in ³¹P MRS spectra due to overlapping signals. Detailed procedures for quantifying the pre-existing correlations between ³¹P-containing metabolites are presented to facilitate incorporation of spectral correlations into statistical modeling in clinical correlation studies.

Monte Carlo study of metabolite correlations originating from spectral overlap

Monte Carlo simulations and a mathematical model of spectral fitting were used to study the correlations between metabolites with overlapping resonances. The dependence of the polarity and the magnitude of cross-correlation coefficients between overlapping metabolites on the spectral patterns of MRS signals was investigated. The results demonstrate the importance of quantifying metabolite correlations originating from spectral overlap as they may confound determination of correlations of biological origin. The findings also indicate that it is possible to minimize unwanted metabolite correlations by altering spectral patterns in the presence of significant spectral overlap.

Single-Voxel Proton Magnetic Resonance Spectroscopy of the Thalamus in Idiopathic Epileptic Dogs and in Healthy Control Dogs

The role of magnetic resonance spectroscopy (MRS) in the investigation of brain metabolites in epileptic syndromes in dogs has not been explored systematically to date. The aim of this study was to investigate metabolites in the thalamus in dogs affected by idiopathic epilepsy (IE) with and without antiepileptic drug treatment (AEDT) and to compare them to unaffected controls. Our hypothesis is that similar to humans with generalized epilepsy and loss of consciousness, N-acetyl aspartate (NAA) would be reduced, and glutamate–glutamine (Glx) would be increased in treated and untreated IE in comparison with the control group. In this prospective case–control study, Border Collie (BC) and Greater Swiss Mountain dog (GSMD) were divided into three groups: (1) healthy controls, IE with generalized tonic–clonic seizures with (2) and without (3) AEDT. A total of 41 BC and GSMD were included using 3 Tesla single-voxel proton MRS of the thalamus (PRESS localization, shortest TE, TR = 2000 ms, NSA = 240). After exclusion of 11 dogs, 30 dogs (18 IE and 12 healthy controls) remained available for analysis. Metabolite concentrations were estimated with LCModel using creatine as reference and compared using Kruskal–Wallis and Wilcoxon rank-sum tests. The Kruskal–Wallis test revealed significant differences in the NAA-to-creatine (p = 0.04) and Glx-to-creatine (p = 0.03) ratios between the three groups. The Wilcoxon rank-sum test further showed significant reduction in the NAA/creatine ratio in idiopathic epileptic dogs under AEDT compared to epileptic dogs without AEDT (p = 0.03) and compared to healthy controls (p = 0.03). In opposite to humans, Glx/creatine ratio was significantly reduced in dogs with IE under AEDT compared to epileptic dogs without AEDT (p = 0.03) and controls (p = 0.02). IE without AEDT and healthy controls did not show significant difference, neither in NAA/creatine (p = 0.60), nor in Glx-to-creatine (p = 0.55) ratio. In conclusion, MRS showed changes in dogs with IE and generalized seizures under AEDT, but not in those without AEDT. Based upon these results, MRS can be considered a useful advanced imaging technique for the evaluation of dogs with IE in the clinical and research settings.

ProFit-1D-A 1D fitting software and open-source validation data sets

PURPOSE

Accurate and precise MRS fitting is crucial for metabolite concentration quantification of ¹ H-MRS spectra. LCModel, a spectral fitting software, has shown to have certain limitations to perform advanced spectral fitting by previous literature. Herein, we propose an open-source spectral fitting algorithm with adaptive spectral baseline determination and more complex cost functions.

THEORY

The MRS spectra are characterized by several parameters, which reflect the environment of the contributing metabolites, properties of the acquisition sequence, or additional disturbances. Fitting parameters should accurately describe these parameters. Baselines are also a major contributor to MRS spectra, in which smoothness of the spline baselines used for fitting can be adjusted based on the properties of the spectra. Three different cost functions used for the minimization problem were also investigated.

METHODS

The newly developed ProFit-1D fitting algorithm is systematically evaluated for simulations of several types of possible in vivo parameter variations. Although accuracy and precision are tested with simulated spectra, spectra measured in vivo at 9.4 T are used for testing precision using subsets of averages. ProFit-1D fitting results are also compared with LCModel.

RESULTS

Both ProFit-1D and LCModel fitted the spectra well with induced parameter and baseline variations. ProFit-1D proved to be more accurate than LCModel for simulated spectra. However, LCModel showed a somewhat increased precision for some spectral simulations and for in vivo data.

CONCLUSION

The open-source ProFit-1D fitting algorithm demonstrated high accuracy while maintaining precise metabolite concentration quantification. Finally, through the newly proposed cost functions, new ways to improve fitting were shown.

Comparison of different linear-combination modeling algorithms for short-TE proton spectra

Short-TE proton MRS is used to study metabolism in the human brain. Common analysis methods model the data as a linear combination of metabolite basis spectra. This large-scale multi-site study compares the levels of the four major metabolite complexes in short-TE spectra estimated by three linear-combination modeling (LCM) algorithms. 277 medial parietal lobe short-TE PRESS spectra (TE = 35 ms) from a recent 3 T multi-site study were preprocessed with the Osprey software. The resulting spectra were modeled with Osprey, Tarquin and LCModel, using the same three vendor-specific basis sets (GE, Philips and Siemens) for each algorithm. Levels of total N-acetylaspartate (tNAA), total choline (tCho), myo-inositol (mI) and glutamate + glutamine (Glx) were quantified with respect to total creatine (tCr). Group means and coefficient of variations of metabolite estimates agreed well for tNAA and tCho across vendors and algorithms, but substantially less so for Glx and mI, with mI systematically estimated as lower by Tarquin. The cohort mean coefficient of determination for all pairs of LCM algorithms across all datasets and metabolites was R 2 ¯ = 0.39, indicating generally only moderate agreement of individual metabolite estimates between algorithms. There was a significant correlation between local baseline amplitude and metabolite estimates (cohort mean R 2 ¯ = 0.10). While mean estimates of major metabolite complexes broadly agree between linear-combination modeling algorithms at group level, correlations between algorithms are only weak-to-moderate, despite standardized preprocessing, a large sample of young, healthy and cooperative subjects, and high spectral quality. These findings raise concerns about the comparability of MRS studies, which typically use one LCM software and much smaller sample sizes.

Cerebral phosphoester signals measured by 31P magnetic resonance spectroscopy at 3 and 7 Tesla

Abnormal cell membrane metabolism is associated with many neuropsychiatric disorders. Free phosphomonoesters and phosphodiesters, which can be detected by in vivo 31P magnetic resonance spectroscopy (MRS), are important cell membrane building blocks. However, the quantification of phosphoesters has been highly controversial even in healthy individuals due to overlapping signals from macromolecule membrane phospholipids (MP). In this study, high signal-to-noise ratio (SNR) cerebral 31P MRS spectra were acquired from healthy volunteers at both 3 and 7 Tesla. Our results indicated that, with minimal spectral interference from MP, the [phosphocreatine (PCr)]/[phosphocholine (PC) + glycerophosphocholine (GPC)] ratio measured at 7 Tesla agreed with its value expected from biochemical constraints. In contrast, the 3 Tesla [PCr]/[PC+GPC] ratio obtained using standard spectral fitting procedures was markedly smaller than the 7 Tesla ratio and than the expected value. The analysis suggests that the commonly used spectral model for MP may fail to capture its complex spectral features at 3 Tesla, and that additional prior knowledge is necessary to reliably quantify the phosphoester signals at low magnetic field strengths when spectral overlapping is significant.

PCA denoising and Wiener deconvolution of 31 P 3D CSI data to enhance effective SNR and improve point spread function

Purpose

This study evaluates the performance of 2 processing methods, that is, principal component analysis‐based denoising and Wiener deconvolution, to enhance the quality of phosphorus 3D chemical shift imaging data.

Methods

Principal component analysis‐based denoising increases the SNR while maintaining spectral information. Wiener deconvolution reduces the FWHM of the voxel point spread function, which is increased by Hamming filtering or Hamming‐weighted acquisition. The proposed methods are evaluated using simulated and in vivo 3D phosphorus chemical shift imaging data by 1) visual inspection of the spatial signal distribution; 2) SNR calculation of the PCr peak; and 3) fitting of metabolite basis functions.

Results

With the optimal order of processing steps, we show that the effective SNR of in vivo phosphorus 3D chemical shift imaging data can be increased. In simulations, we show we can preserve phosphorus‐containing metabolite peaks that had an SNR < 1 before denoising. Furthermore, using Wiener deconvolution, we were able to reduce the FWHM of the voxel point spread function with only partially reintroducing Gibb‐ringing artifacts while maintaining the SNR. After data processing, fitting of the phosphorus‐containing metabolite signals improved.

Conclusion

In this study, we have shown that principal component analysis‐based denoising in combination with regularized Wiener deconvolution allows increasing the effective spectral SNR of in vivo phosphorus 3D chemical shift imaging data, with reduction of the FWHM of the voxel point spread function. Processing increased the effective SNR by at least threefold compared to Hamming weighted acquired data and minimized voxel bleeding. With these methods, fitting of metabolite amplitudes became more robust with decreased fitting residuals.

Local and Interregional Neurochemical Associations Measured by Magnetic Resonance Spectroscopy for Studying Brain Functions and Psychiatric Disorders

Magnetic resonance spectroscopy (MRS) studies have found significant correlations among neurometabolites (e.g., between glutamate and GABA) across individual subjects and altered correlations in neuropsychiatric disorders. In this article, we discuss neurochemical associations among several major neurometabolites which underpin these observations by MRS. We also illustrate the role of spectral editing in eliminating unwanted correlations caused by spectral overlapping. Finally, we describe the prospects of mapping macroscopic neurochemical associations across the brain and characterizing excitation-inhibition balance of neural networks using glutamate- and GABA-editing MRS imaging.