Abstract
Background
Quantification of metabolites concentrations in institutional unit (IU) is important for inter-subject and long-term comparisons in the applications of magnetic resonance spectroscopy (MRS). Recently, deep learning (DL) algorithms have found a variety of applications on the process of MRS data. A quantification strategy compatible to DL base MRS spectral processing method is, therefore, useful.
Materials and methods
This study aims to investigate whether metabolite concentrations quantified using a convolutional neural network (CNN) based method, coupled with a scaling procedure that normalizes spectral signals for CNN input and linear regression, can effectively reflect variations in metabolite concentrations in IU across different brain regions with varying signal-to-noise ratios (SNR) and linewidths (LW). An error index based on standard error (SE) is proposed to indicate the confidence levels associated with metabolite predictions. In vivo MRS spectra were acquired from three brain regions of 43 subjects using a 3T system.
Results
The metabolite concentrations in IU of five major metabolites, quantified using CNN and LCModel, exhibit similar ranges with Pearson’s correlation coefficients ranging from 0.24 to 0.78. The SE of the metabolites shows a positive correlation with Cramer-Rao lower bound (CRLB) (r=0.46) and absolute CRLB (r=0.81), calculated by multiplying CRLBs with the quantified metabolite content.
Conclusion
In conclusion, the CNN based method with the proposed scaling procedures can be employed to quantify in vivo MRS spectra and derive metabolites concentrations in IU. The SE can be used as error index, indicating predicted uncertainties for metabolites and sharing information similar to the absolute CRLB.
Similar content being viewed by others
Data availability
The data cannot be shared at this time as the data also forms part of an ongoing study. Please contact the corrosponding author for futher access.
References
Niddam DM, Lai KL, Tsai SY, Lin YR, Chen WT, Fuh JL, Wang SJ (2020) Brain metabolites in chronic migraine patients with medication overuse headache. Cephalalgia 40(8):851–862
Niddam DM, Wang SJ, Tsai SY (2021) Pain sensitivity and the primary sensorimotor cortices: a multimodal neuroimaging study. Pain 162(3):846–855
Niddam DM, Lai KL, Tsai SY, Lin YR, Chen WT, Fuh JL, Wang SJ (2018) Neurochemical changes in the medial wall of the brain in chronic migraine. Brain 141:377–390
Schur RR, Draisma LW, Wijnen JP, Boks MP, Koevoets MG, Joels M, Klomp DW, Kahn RS, Vinkers CH (2016) Brain GABA levels across psychiatric disorders: a systematic literature review and meta-analysis of (1) H-MRS studies. Hum Brain Mapp 37(9):3337–3352
Birch R, Peet AC, Dehghani H, Wilson M (2017) Influence of macromolecule baseline on (1) H MR spectroscopic imaging reproducibility. Magn Reson Med 77(1):34–43
Tsai SY, Lin YR, Lin HY, Lin FH (2019) Reduction of lipid contamination in MR spectroscopy imaging using signal space projection. Magn Reson Med 81(3):1486–1498
Tkac I, Deelchand D, Dreher W, Hetherington H, Kreis R, Kumaragamage C, Povazan M, Spielman DM, Strasser B, de Graaf RA (2021) Water and lipid suppression techniques for advanced (1) H MRS and MRSI of the human brain: experts’ consensus recommendations. NMR Biomed 34(5):e4459
Jiru F, Skoch A, Wagnerova D, Dezortova M, Hajek M (2013) jSIPRO - analysis tool for magnetic resonance spectroscopic imaging. Comput Methods Programs Biomed 112(1):173–188
Borbath T, Murali-Manohar S, Dorst J, Wright AM, Henning A (2021) ProFit-1D-A 1D fitting software and open-source validation data sets. Magn Reson Med 86(6):2910–2929
Near J, Harris AD, Juchem C, Kreis R, Marjanska M, Oz G, Slotboom J, Wilson M, Gasparovic C (2021) Preprocessing, analysis and quantification in single-voxel magnetic resonance spectroscopy: experts’ consensus recommendations. NMR Biomed 34(5):e4257
Edden RA, Puts NA, Harris AD, Barker PB, Evans CJ (2014) Gannet: A batch-processing tool for the quantitative analysis of gamma-aminobutyric acid-edited MR spectroscopy spectra. J Magn Reson Imaging 40(6):1445–1452
Simpson R, Devenyi GA, Jezzard P, Hennessy TJ, Near J (2017) Advanced processing and simulation of MRS data using the FID appliance (FID-A)-An open source, MATLAB-based toolkit. Magn Reson Med 77(1):23–33
Provencher SW (2001) Automatic quantitation of localized in vivo 1H spectra with LCModel. NMR Biomed 14(4):260–264
Jablonski M, Starcukova J, Starcuk Z Jr (2017) Processing tracking in jMRUI software for magnetic resonance spectra quantitation reproducibility assurance. BMC Bioinformatics 18(1):56
Wilson M, Reynolds G, Kauppinen RA, Arvanitis TN, Peet AC (2011) A constrained least-squares approach to the automated quantitation of in vivo (1)H magnetic resonance spectroscopy data. Magn Reson Med 65(1):1–12
Chong DG, Kreis R, Bolliger CS, Boesch C, Slotboom J (2011) Two-dimensional linear-combination model fitting of magnetic resonance spectra to define the macromolecule baseline using FiTAID, a Fitting Tool for Arrays of Interrelated Datasets. MAGMA 24(3):147–164
Maudsley AA, Domenig C, Govind V, Darkazanli A, Studholme C, Arheart K, Bloomer C (2009) Mapping of brain metabolite distributions by volumetric proton MR spectroscopic imaging (MRSI). Magn Reson Med 61(3):548–559
Gasparovic C, Song T, Devier D, Bockholt HJ, Caprihan A, Mullins PG, Posse S, Jung RE, Morrison LA (2006) Use of tissue water as a concentration reference for proton spectroscopic imaging. Magn Reson Med 55(6):1219–1226
Lecocq A, Le Fur Y, Amadon A, Vignaud A, Cozzone PJ, Guye M, Ranjeva JP (2015) Fast water concentration mapping to normalize (1)H MR spectroscopic imaging. MAGMA 28(1):87–100
Rizzo R, Dziadosz M, Kyathanahally SP, Shamaei A, Kreis R (2023) Quantification of MR spectra by deep learning in an idealized setting: Investigation of forms of input, network architectures, optimization by ensembles of networks, and training bias. Magn Reson Med 89(5):1707–1727
Kreis R (2016) The trouble with quality filtering based on relative Cramer-Rao lower bounds. Magn Reson Med 75(1):15–18
Kyathanahally SP, Doring A, Kreis R (2018) Deep learning approaches for detection and removal of ghosting artifacts in MR spectroscopy. Magn Reson Med 80(3):851–863
Gurbani SS, Schreibmann E, Maudsley AA, Cordova JS, Soher BJ, Poptani H, Verma G, Barker PB, Shim H, Cooper LAD (2018) A convolutional neural network to filter artifacts in spectroscopic MRI. Magn Reson Med 80(5):1765–1775
Gurbani SS, Sheriff S, Maudsley AA, Shim H, Cooper LAD (2019) Incorporation of a spectral model in a convolutional neural network for accelerated spectral fitting. Magn Reson Med 81(5):3346–3357
Shamaei A, Starcukova J, Starcuk Z Jr (2023) Physics-informed deep learning approach to quantification of human brain metabolites from magnetic resonance spectroscopy data. Comput Biol Med 158:106837
Lee H, Lee HH, Kim H (2020) Reconstruction of spectra from truncated free induction decays by deep learning in proton magnetic resonance spectroscopy. Magn Reson Med 84(2):559–568
Lee HH, Kim H (2019) Intact metabolite spectrum mining by deep learning in proton magnetic resonance spectroscopy of the brain. Magn Reson Med 82(1):33–48
Lee HH, Kim H (2020) Deep learning-based target metabolite isolation and big data-driven measurement uncertainty estimation in proton magnetic resonance spectroscopy of the brain. Magn Reson Med 84(4):1689–1706
Govindaraju V, Young K, Maudsley AA (2000) Proton NMR chemical shifts and coupling constants for brain metabolites. NMR Biomed 13(3):129–153
Opstad KS, Bell BA, Griffiths JR, Howe FA (2008) Toward accurate quantification of metabolites, lipids, and macromolecules in HRMAS spectra of human brain tumor biopsies using LCModel. Magn Reson Med 60(5):1237–1242
Peng PY, Tsai SY, Lin YR (2016) Toolbox for automatic localization of volume of interest in MRS (ALLVOI). Paper presented at the Proceedings of the 24th Annual Meeting of ISMRM, Singapore
Park YW, Deelchand DK, Joers JM, Hanna B, Berrington A, Gillen JS, Kantarci K, Soher BJ, Barker PB, Park H, Oz G, Lenglet C (2018) AutoVOI: real-time automatic prescription of volume-of-interest for single voxel spectroscopy. Magn Reson Med 80(5):1787–1798
Tsai SY, Fang CH, Wu TY, Lin YR (2016) Effects of frequency drift on the quantification of gamma-aminobutyric acid using MEGA-PRESS. Sci Rep 6:24564
Oz G, Deelchand DK, Wijnen JP, Mlynarik V, Xin L, Mekle R, Noeske R, Scheenen TWJ, Tkac I, Experts' Working Group on Advanced Single Voxel HM (2020) Advanced single voxel(1) H magnetic resonance spectroscopy techniques in humans: experts' consensus recommendations. NMR Biomed. https://doi.org/10.1002/nbm.4236:e4236
Dhamala E, Abdelkefi I, Nguyen M, Hennessy TJ, Nadeau H, Near J (2019) Validation of in vivo MRS measures of metabolite concentrations in the human brain. NMR Biomed 32(3):e4058
Bednarik P, Moheet A, Deelchand DK, Emir UE, Eberly LE, Bares M, Seaquist ER, Oz G (2015) Feasibility and reproducibility of neurochemical profile quantification in the human hippocampus at 3 T. NMR Biomed 28(6):685–693
Chiu PW, Mak HK, Yau KK, Chan Q, Chang RC, Chu LW (2014) Metabolic changes in the anterior and posterior cingulate cortices of the normal aging brain: proton magnetic resonance spectroscopy study at 3 T. Age (Dordr) 36(1):251–264
Lee HH, Kim H (2022) Bayesian deep learning-based (1) H-MRS of the brain: metabolite quantification with uncertainty estimation using Monte Carlo dropout. Magn Reson Med 88(1):38–52
Dziadosz M, Rizzo R, Kyathanahally SP, Kreis R (2023) Denoising single MR spectra by deep learning: miracle or mirage? Magn Reson Med. https://doi.org/10.1002/mrm.29762
Giapitzakis IA, Borbath T, Murali-Manohar S, Avdievich N, Henning A (2019) Investigation of the influence of macromolecules and spline baseline in the fitting model of human brain spectra at 9.4T. Magn Reson Med 81(2):746–758
Marjanska M, Terpstra M (2021) Influence of fitting approaches in LCModel on MRS quantification focusing on age-specific macromolecules and the spline baseline. NMR Biomed 34(5):e4197
Acknowledgements
The authors thank the Taiwan Mind & Brain Imaging Center (TMBIC) and National Chengchi University for the instrument availability. The TMBIC is supported by the National Science and Technology Council, Taiwan. This work was supported in part by grants from the National Science and Technology Council, Taiwan (108-2314-B-004-001, 111-2314-B-004 -001, 107-2221-E-011-053)
Author information
Authors and Affiliations
Contributions
S.-Y. T. conceived and designed the experiments. Y.-L. H. and S.-Y. T. performed the experiments. Y.-L. H. and Y.-R. L. analyzed the data. S.-Y. T, Y.-L. H., and Y.-R. L interpreted the data. S.-Y. T. and Y.-L. H. wrote the main manuscript text and prepared the tables and figures. S.-Y. T. and Y.-R. L revised the manuscript critically for important intellectual content. All authors reviewed the manuscript and approved the final version.
Corresponding author
Ethics declarations
Conflict of interest
All authors have no conflict of interest to declare.
Ethical standard
This is a prospective institutional review board approved study.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Huang, YL., Lin, YR. & Tsai, SY. Comparison of convolutional-neural-networks-based method and LCModel on the quantification of in vivo magnetic resonance spectroscopy. Magn Reson Mater Phy (2023). https://doi.org/10.1007/s10334-023-01120-z
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s10334-023-01120-z