Metabolites

2023-08-27 13:41| 来源: 网络整理| 查看: 265

Next Article in Journal Metabolomic Footprint of Disrupted Energetics and Amino Acid Metabolism in Neurodegenerative Diseases: Perspectives for Early Diagnosis and Monitoring of Therapy Previous Article in Journal Assessing the Effectiveness of Chemical Marker Extraction from Amazonian Plant Cupuassu (Theobroma grandiflorum) by PSI-HRMS/MS and LC-HRMS/MS Journals Active Journals Find a Journal Proceedings Series Topics Information For Authors For Reviewers For Editors For Librarians For Publishers For Societies For Conference Organizers Open Access Policy Institutional Open Access Program Special Issues Guidelines Editorial Process Research and Publication Ethics Article Processing Charges Awards Testimonials Author Services Initiatives Sciforum MDPI Books Preprints.org Scilit SciProfiles Encyclopedia JAMS Proceedings Series About Overview Contact Careers News Press Blog Sign In / Sign Up Notice clear Notice

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

Continue Cancel clear

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess.

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Journals Active Journals Find a Journal Proceedings Series Topics Information For Authors For Reviewers For Editors For Librarians For Publishers For Societies For Conference Organizers Open Access Policy Institutional Open Access Program Special Issues Guidelines Editorial Process Research and Publication Ethics Article Processing Charges Awards Testimonials Author Services Initiatives Sciforum MDPI Books Preprints.org Scilit SciProfiles Encyclopedia JAMS Proceedings Series About Overview Contact Careers News Press Blog Sign In / Sign Up Submit Journals Metabolites Volume 13 Issue 3 10.3390/metabo13030368

Submit to this Journal Review for this Journal Edit a Special Issue ► ▼ Article Menu Article Menu Academic Editor

Andreas Stadlbauer Subscribe SciFeed Recommended Articles Related Info Links PubMed/Medline Google Scholar More by Author Links on DOAJ Baek, H. on Google Scholar Baek, H. on PubMed Baek, H. /ajax/scifeed/subscribe Article Views Citations - Table of Contents Altmetric share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up ... Endorse textsms ... Comment Need Help? Support

Find support for a specific problem in the support section of our website.

Get Support Feedback

Please let us know what you think of our products and services.

Give Feedback Information

Visit our dedicated information section to learn more about MDPI.

Get Information clear JSmol Viewer clear first_page settings Order Article Reprints Font Type: Arial Georgia Verdana Font Size: Aa Aa Aa Line Spacing:    Column Width:    Background: Open AccessArticle Experimental Basis Sets of Quantification of Brain 1H-Magnetic Resonance Spectroscopy at 3.0 T by

Hyeon-Man Baek

Hyeon-Man Baek Scilit Preprints.org Google Scholar 1,2 1 Department of Health Sciences and Technology, GAIHST, Gachon University, Incheon 21999, Republic of Korea 2 Department of Molecular Medicine, Lee Gil Ya Cancer and Diabetes Institute, Gachon University, Incheon 21999, Republic of Korea Metabolites 2023, 13(3), 368; https://doi.org/10.3390/metabo13030368 Received: 14 January 2023 / Revised: 9 February 2023 / Accepted: 27 February 2023 / Published: 1 March 2023 (This article belongs to the Topic Metabolism and Health) Download Download PDF Download PDF with Cover Download XML Download Epub Browse Figures Review Reports Versions Notes

Abstract: In vivo short echo time (TE) proton magnetic resonance spectroscopy (1H-MRS) is a useful method for the quantification of human brain metabolites. The purpose of this study was to evaluate the performance of an in-house, experimentally measured basis set and compare it with the performance of a vendor-provided basis set. A 3T clinical scanner with 32-channel receive-only phased array head coil was used to generate 16 brain metabolites for the metabolite basis set. For voxel localization, point-resolved spin-echo sequence (PRESS) was used with volume of interest (VOI) positioned at the center of the phantoms. Two different basis sets were subjected to linear combination of model spectra of metabolite solutions in vitro (LCModel) analysis to evaluate the in-house acquired in vivo 1H-MR spectra from the left prefrontal cortex of 22 healthy subjects. To evaluate the performance of the two basis sets, the Cramer-Rao lower bounds (CRLBs) of each basis set were compared. The LCModel quantified the following metabolites and macromolecules: alanine (Ala), aspartate (Asp), γ-amino butyric acid (GABA), glucose (Glc), glutamine (Gln), glutamate (Glu), glutathione (GHS), Ins (myo-Inositol), lactate (Lac), N-acetylaspartate (NAA), N-acetylaspartylglutamate (NAAG), taurine (Tau), phosphoryl-choline + glycerol-phosphoryl-choline (tCho), N-acetylaspartate + N-acetylaspartylglutamate (tNA), creatine + phosphocreatine (tCr), Glu + Gln (Glx) and Lip13a, Lip13b, Lip09, MM09, Lip20, MM20, MM12, MM14, MM17, Lip13a + Lip13b, MM14 + Lip13a + Lip13b + MM12, MM09 + Lip09, MM20 + Lip20. Statistical analysis showed significantly different CRLBs: Asp, GABA, Gln, GSH, Ins, Lac, NAA, NAAG, Tau, tCho, tNA, Glx, MM20, MM20 + Lip20 (p < 0.001), tCr, MM12, MM17 (p < 0.01), and Lip20 (p < 0.05). The estimated ratio of cerebrospinal fluid (CSF) in the region of interest was calculated to be about 5%. Fitting performances are better, for the most part, with the in-house basis set, which is more precise than the vendor-provided basis set. In particular, Asp is expected to have reliable CRLB ( 10). The horizontal axis represents the FWHM of the water peak before water suppression, and the vertical axis represents the SNR after water suppression. SNR is the maximum signal intensity in the 0–3.4 ppm region, with respect to the standard deviation of the noise in the 9–11 ppm range. The FWHMs of 22 spectra were in the 8-Hz range. Therefore, shimming in the VOI was considered successful. If the shimming procedure is not well-performed, the width of the peaks in the spectra is widened, resulting in a greater overlap of the metabolite peaks, which makes distinguishing metabolite peaks more difficult. It can also result in poor fitting accuracy and eventually make the quantification of metabolite concentrations more difficult. Results from 22 spectra revealed that all measured SNRs were greater than 10. This indicates that peaks for metabolites such as NAA, tCr, and tCho are 10 times greater in magnitude than the noise. After zero-order phase correction, water peak heights before and after water suppression were measured. Results were 0.7 ± 0.4 (mean ± SD) %, less than 1%, indicating good water suppression. 3.3. LCModel Spectral Fitting AnalysisFigure 3 shows the mean spectrum of 22 spectra. Figure 3a shows a spectral analysis result using the basis set obtained experimentally on the same system. Figure 3b shows a spectral analysis result using the vendor provided basis set, which was not generated from the experimental basis set and in vivo data. The black line in Figure 3 indicates phased in vivo data, and the red line indicates fitted data. The lines below the red line show the baseline and each fitted metabolite signal. The blue line at the top of the figure represents the fitting residues of the difference between in vivo and fitted data. The Shapiro–Wilk normality test of fitting residues showed that using the in-house experimental basis set (0 of 22, p < 0.05) was better than using the provided basis set fitting (6 of 22, p < 0.05).Table 1 shows the mean CRLBs and standard errors of each of the metabolites and macromolecules. The result indicates that fitting precisions using the in-house experimental basis set were better than the provided simulated basis set for Asp, GABA, Gln, GSH, Ins, Lac, NAA, NAAG, Tau, tCho, tNA, Glx (p < 0.001), and MM12 (p < 0.01). On the other hand, there were opposite results for tCr (p < 0.01), Lip20 (p < 0.05), MM20 (p < 0.001), MM17 (p < 0.01), and MM20 + Lip20 (p < 0.001), and there were no precision differences for Ala, Glc, Glu, Lip13a, Lip13b, Lip09, MM09, MM14, Lip13a + Lip13b, MM14 + Lip13a + Lip13b + MM12, and MM09 + Lip09.Table 2 shows the mean concentrations and standard errors of each of the metabolites and macromolecules. The paired t-test indicates that quantified concentrations using the in-house experimental basis set are significantly lower than the provided simulated basis set for GABA, Gln, GSH, Ins, Lac, NAA, NAAG, Tau, tCho, tNA, Lip09, and MM17 (p < 0.001). The opposite results were found for Asp, Glu, Glx, tCr, Lip20, MM20, MM12, MM20 + lip20 (p < 0.001), MM09, and MM14 + Lip13a + Lip13b + MM12 (p < 0.05). Additionally, there were no significant differences for Ala, Glc, Lip13a, Lip13b, MM14, Lip13a + Lip13b, and MM09 + Lip09. 3.4. Correlation MatricesFigure 4 shows the correlation matrices of metabolite and macromolecules concentrations, as quantified using both basis sets basis. The blue colors indicate positive correlations, red colors indicate negative correlations, and blanks indicate no significant correlations. Before thresholding, Figure 4a (left, the in-house experimental basis set) and Figure 4b (right, the provided simulated basis set) showed very similar patterns, and we could not find any noticeable differences between the two matrices; thus, we adjusted significant level of the P value to 0.1. There were negative correlations in the left matrix between GABA and MM09, GABA and MM12, Glc and NAA, Glc and tNA, Glc and Lip09, Gln and MM17, Lac and Lip13b, and Lac and Lip13a + Lip13b. On the other hand, there were inverse correlations in the right matrix between Asp and NAAG (p = 0.051), Gln and MM14, Gln and MM17, Lac and Lip13a, Lac and Lip13b, Lac and Lip13a + Lip13b, Lac and MM14 + Lip13a + Lip13b + MM12, tCr and Lip13b. 3.5. In Vitro 1H-MRS Phantom StudyTable 3 shows a quantification result of standard MRS phantom. Metabolite concentrations were underestimated from true concentrations (Ins; -1.4 mM, Lac; -1.63 mM, Cho; −0.55 mM, NAA; −0.92 mM, Cr; −0.15 mM, and Glu; −0.15 mM) using the in-house experimental basis set. On the other hand, concentrations were overestimated (Ins; +0.39 mM, Cho; +0.54 mM, NAA; +0.83 mM, and Cr; +0.37 mM) and underestimated (Lac; −0.67 mM, and Glu; −0.74 mM). We could not find any notable differences in relative units. However, in an absolute unit, we could find that the experimental basis set was more precise for Ins, Lac, Cho, NAA, and Glu, but not for Cr, than the provided simulated basis set. 3.6. Coefficient of VariationTo investigate the different variations of the results using the two different approaches, coefficients of variation (CV) were compared. CV is defined as metabolite concentration divided by its standard deviation. Figure 5 shows CV versus mean CRLBs for the two approaches. We use CRLBs in relative unit (i.e., %SD) to compare with the study and to interpret the results. We included all subjects’ data. CVs of metabolites for which CRLBs are less than 15%: Glu (10.1 vs. 7.6%, the in-house experimental basis set vs. the provided simulated basis set), Ins (12.4 vs. 12.3%), NAA (18.7 vs. 7.2%), tCho (20.5 vs. 15.5%), tNA (14.1 vs. 20.1%), tCr (18.1 vs. 16.4%), and Glx (7.4 vs. 9.0%). The metabolites appeared above the identity line in Figure 5b. The CVs for which CRLBs are less than 35% include: Gln (26.2 vs. 25.0%), GSH (12.3 vs. 12.7%), MM09 (10.7 vs. 9.9%), MM20 (9.4 vs. 9.6%), MM12 (11.8 vs. 11.3%), MM14 (15.5 vs. 11.9%), MM17 (12.6 vs. 19.4%), MM14 + Lip13a + Lip13b + MM12 (10.7 vs. 9.3%), MM09 + Lip09 (6.8 vs. 7.2%), and MM20 + Lip20 (10.0 vs. 10.5%). The metabolites and macromolecules appear below the identity line in Figure 5b. The CVs for which CRLBs are greater than 35% include: GABA (125.5 vs. 47%), NAAG (192.6 vs. 146.2%), and Lac (141.6 vs. 134.2%). The metabolites appear above the identity line in Figure 5a. However, Asp (25.0 vs. 155.2%) and Tau (49.5 vs. 28.6%) appeared in different regions from each other. 4. DiscussionIn this study, we aimed to evaluate the performance of the in-house experimental basis set, compared to the vendor, provided simulated basis set by applying in-house acquired in vivo 1H-MRS data. Data were acquired from the left prefrontal cortex of 22 normal subjects. Fitting errors from two approaches using CRLBs were compared. We found that the LCModel quantification with the in-house experimental basis set had significantly lower CRLBs than the provided simulated basis set for most metabolites and macromolecules at 3T.According to Wilson et al. [16], there are two main reasons for using different basis sets. First, metabolite signals can be different for each basis set for several reasons, including signal amplitude, SNR, or baseline. This type of difference is referred to as differences in basis sets. Second, the fitting algorithms may cover different solutions. In this case, the fitting residue can be similar, but with different quantification results. For example, high PC and low GPC concentrations may yield similar results as low PC and high GPC. These differences are referred to as fitting instability. When using the two different basis sets, fitting instabilities are readily apparent. Therefore, to measure the difference between basis sets accurately, it is important to reduce the fitting algorithm-dependent fitting instabilities as much as possible. Discrepancy due to differences in basis sets should be independent of the quality of data. Wilson et al. [16] suggested that a level of agreement of ±0.99 mM between simulated and experimental basis sets is sufficiently small for them to be used interchangeably. While small fitting instabilities are inevitable, it should not be noticeable for high quality data. They reported that fitting instabilities are dominant in poor quality data because the poor quality results in a loss of intrinsic information.Graaf et al. [20] made protocols for spectrum quality assessment and applied it to their study. The protocol contains investigations of FWHM (10), and artifacts. The thresholds of FWHM and SNR for quality assessment were determined by experts. This protocol was more conservative than the criteria determined by a supervised pattern recognition method. In our study, the results satisfied the above-mentioned protocol (e.g., FWHM < 8 Hz, SNR > 10, no artifact). The difference is B0 field strength. The Graaf et al. [20] study was performed at 1.5T, while we experimented at 3T. The increased strength of the magnetic field linearly correlates with SNR; our results show a larger SNR value using a 3T magnetic field than the SNR value of 20 from the experiment that used a 1.5T magnetic field. On the other hand, if the field strength is increased, the local field inhomogeneity is increased, and it makes T2 relaxation time shorter. Therefore, the line width of spectral peaks is broadened, and FWHM is increased. If SNR is low, it is difficult to distinguish between metabolite peaks and noise, and fitting instabilities will increase, making the quantification of metabolite concentrations difficult. If water suppression is not adequate, water peaks hinder the fitting of other metabolites, making quantification difficult. In our study, FWHM, SNR, and the extent of water suppression show that these parameters do not have a significant impact on quantitative analysis (Figure 2).Differences between the fitting errors of the two approaches can be due to the following reasons. The RF pulse shapes used to acquire the two basis sets are different for the two approaches. The basis sets used in this study were in-house experimentally generated and received from a vendor. Both basis sets were acquired using the PRESS sequence, and TE was 35 ms. The in-house experimental basis set was generated using the same parameters used in in vivo and standard MRS phantom experiments. A pulse diagram contains RF pulses (excitation and refocusing) that were taken in real time. The pulse diagram also contains various gradients (slice selection and spoiler) that have finite ramp times. MRI machine might have had delayed times, turning on the pulses resulting in an imperfect shimming. Therefore, magnetic field inhomogeneity during real experiments is inherent. On the other hand, the provided simulated basis set assumes ideal hard pulses and does not consider localization, pulse offsets, pulse power, and pulse lengths, which results in two different basis sets.Kaiser et al. [29] showed a comparison of ideal and 3D-localised PRESS simulations with experimental 3D PRESS-localized phantom spectra of Lac, NAA, Glu, and Ins at different field strengths. They reported that PRESS-localized spectra contained discrepancies between ideal and localized conditions. They also demonstrated that those discrepancies are more severe in higher fields. As such, the difference in the performance of two basis sets and the discrepancies in spectral line shape may be due to the inconsistency of localization conditions. This difference is expected to be more apparent for the strongly coupled metabolites because of their complex appearances. Compared to weakly coupled or non-coupled metabolites, strongly coupled metabolites can result in a more noticeable difference [30]. We investigated regularization parameters (e.g., α S and α B in the LCModel method (Provencher [21])) to identify the localization effects from our data. The output parameters α S (8.4 ± 19.08 vs. 62.5 ± 77.0, mean ± SD), α B (0.065 ± 0.02 vs. 0.073 ± 0.03), and α S / α B (129 ± 283.15 vs. 856 ± 1572.06) were significantly lower with in-house experimental basis set (p < 0.05). Therefore, we believe that fitting with the provided simulated basis set needed more regularization terms to account for line shape discrepancies than with in-house experimental basis set fitting, in order to minimize the objective function.Recently, Deelchand et al. [30] performed a simulation study to investigate how the quantitative accuracy of the metabolite can be increased by using a short echo time when the B0 field increases in the human brain. They reported that greater improvement in quantification precision is obtained for J-coupled metabolites than for singlets as the B0 increases. They also reported that additional improvement of quantification precision is expected for J-coupled metabolites because of relatively simple spectral patterns, which can be easily distinguished by reducing the peak overlap. They suggested the possibility of quantifying metabolites, such as NAA, tCho, Ins, Glx, tCr, and Cr, reliably (CRLB < 25%) at 3T. In our study, we could also quantify GSH, Glu (

【本文地址】

公司简介

联系我们