Intact metabolite spectrum mining by deep learning in proton magnetic resonance spectroscopy of the brain

Hyeong Hun Lee, Hyeonjin Kim

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Purpose: To develop a robust method for brain metabolite quantification in proton magnetic resonance spectroscopy ( 1 H-MRS) using a convolutional neural network (CNN) that maps in vivo brain spectra that are typically degraded by low SNR, line broadening, and spectral baseline into noise-free, line-narrowed, baseline-removed intact metabolite spectra. Methods: A CNN was trained (n = 40 000) and tested (n = 5000) on simulated brain spectra with wide ranges of SNR (6.90–20.74) and linewidth (10–20 Hz). The CNN was further tested on in vivo spectra (n = 40) from five healthy volunteers with substantially different SNR, and the results were compared with those from the LCModel analysis. A Student t test was performed for the comparison. Results: Using the proposed method the mean-absolute-percent-errors (MAPEs) in the estimated metabolite concentrations were 12.49% ± 4.35% for aspartate, creatine (Cr), γ-aminobutyric acid (GABA), glucose, glutamine, glutamate, glutathione (GSH), myo-Inositol (mI), N-acetylaspartate, phosphocreatine (PCr), phosphorylethanolamine, and taurine over the whole simulated spectra in the test set. The metabolite concentrations estimated from in vivo spectra were close to the reported ranges for the proposed method and the LCModel analysis except mI, GSH, and especially Cr/PCr for the LCModel analysis, and phosphorylcholine to glycerophosphorylcholine ratio (PC/GPC) for both methods. The metabolite concentrations estimated across the in vivo spectra with different SNR were less variable with the proposed method (~10% or less) than with the LCModel analysis. Conclusion: The robust performance of the proposed method against low SNR may allow a subminute 1 H-MRS of human brain, which is an important technical development for clinical studies.

Original languageEnglish
Pages (from-to)33-48
Number of pages16
JournalMagnetic Resonance in Medicine
Volume82
Issue number1
DOIs
StatePublished - 1 Jul 2019

Fingerprint

Learning
Brain
Phosphocreatine
Creatine
Inositol
Glycerylphosphorylcholine
Aminobutyrates
Phosphorylcholine
Taurine
Proton Magnetic Resonance Spectroscopy
Glutamine
Aspartic Acid
gamma-Aminobutyric Acid
Glutathione
Noise
Glutamic Acid
Healthy Volunteers
Students
Glucose

Keywords

  • brain
  • convolutional neural network
  • deep learning
  • metabolite quantification
  • proton magnetic resonance spectroscopy

Cite this

@article{4ea8c00a8a064b1c857e1bfe6e147653,
title = "Intact metabolite spectrum mining by deep learning in proton magnetic resonance spectroscopy of the brain",
abstract = "Purpose: To develop a robust method for brain metabolite quantification in proton magnetic resonance spectroscopy ( 1 H-MRS) using a convolutional neural network (CNN) that maps in vivo brain spectra that are typically degraded by low SNR, line broadening, and spectral baseline into noise-free, line-narrowed, baseline-removed intact metabolite spectra. Methods: A CNN was trained (n = 40 000) and tested (n = 5000) on simulated brain spectra with wide ranges of SNR (6.90–20.74) and linewidth (10–20 Hz). The CNN was further tested on in vivo spectra (n = 40) from five healthy volunteers with substantially different SNR, and the results were compared with those from the LCModel analysis. A Student t test was performed for the comparison. Results: Using the proposed method the mean-absolute-percent-errors (MAPEs) in the estimated metabolite concentrations were 12.49{\%} ± 4.35{\%} for aspartate, creatine (Cr), γ-aminobutyric acid (GABA), glucose, glutamine, glutamate, glutathione (GSH), myo-Inositol (mI), N-acetylaspartate, phosphocreatine (PCr), phosphorylethanolamine, and taurine over the whole simulated spectra in the test set. The metabolite concentrations estimated from in vivo spectra were close to the reported ranges for the proposed method and the LCModel analysis except mI, GSH, and especially Cr/PCr for the LCModel analysis, and phosphorylcholine to glycerophosphorylcholine ratio (PC/GPC) for both methods. The metabolite concentrations estimated across the in vivo spectra with different SNR were less variable with the proposed method (~10{\%} or less) than with the LCModel analysis. Conclusion: The robust performance of the proposed method against low SNR may allow a subminute 1 H-MRS of human brain, which is an important technical development for clinical studies.",
keywords = "brain, convolutional neural network, deep learning, metabolite quantification, proton magnetic resonance spectroscopy",
author = "Lee, {Hyeong Hun} and Hyeonjin Kim",
year = "2019",
month = "7",
day = "1",
doi = "10.1002/mrm.27727",
language = "English",
volume = "82",
pages = "33--48",
journal = "Magnetic Resonance in Medicine",
issn = "0740-3194",
publisher = "John Wiley and Sons Inc.",
number = "1",

}

Intact metabolite spectrum mining by deep learning in proton magnetic resonance spectroscopy of the brain. / Lee, Hyeong Hun; Kim, Hyeonjin.

In: Magnetic Resonance in Medicine, Vol. 82, No. 1, 01.07.2019, p. 33-48.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Intact metabolite spectrum mining by deep learning in proton magnetic resonance spectroscopy of the brain

AU - Lee, Hyeong Hun

AU - Kim, Hyeonjin

PY - 2019/7/1

Y1 - 2019/7/1

N2 - Purpose: To develop a robust method for brain metabolite quantification in proton magnetic resonance spectroscopy ( 1 H-MRS) using a convolutional neural network (CNN) that maps in vivo brain spectra that are typically degraded by low SNR, line broadening, and spectral baseline into noise-free, line-narrowed, baseline-removed intact metabolite spectra. Methods: A CNN was trained (n = 40 000) and tested (n = 5000) on simulated brain spectra with wide ranges of SNR (6.90–20.74) and linewidth (10–20 Hz). The CNN was further tested on in vivo spectra (n = 40) from five healthy volunteers with substantially different SNR, and the results were compared with those from the LCModel analysis. A Student t test was performed for the comparison. Results: Using the proposed method the mean-absolute-percent-errors (MAPEs) in the estimated metabolite concentrations were 12.49% ± 4.35% for aspartate, creatine (Cr), γ-aminobutyric acid (GABA), glucose, glutamine, glutamate, glutathione (GSH), myo-Inositol (mI), N-acetylaspartate, phosphocreatine (PCr), phosphorylethanolamine, and taurine over the whole simulated spectra in the test set. The metabolite concentrations estimated from in vivo spectra were close to the reported ranges for the proposed method and the LCModel analysis except mI, GSH, and especially Cr/PCr for the LCModel analysis, and phosphorylcholine to glycerophosphorylcholine ratio (PC/GPC) for both methods. The metabolite concentrations estimated across the in vivo spectra with different SNR were less variable with the proposed method (~10% or less) than with the LCModel analysis. Conclusion: The robust performance of the proposed method against low SNR may allow a subminute 1 H-MRS of human brain, which is an important technical development for clinical studies.

AB - Purpose: To develop a robust method for brain metabolite quantification in proton magnetic resonance spectroscopy ( 1 H-MRS) using a convolutional neural network (CNN) that maps in vivo brain spectra that are typically degraded by low SNR, line broadening, and spectral baseline into noise-free, line-narrowed, baseline-removed intact metabolite spectra. Methods: A CNN was trained (n = 40 000) and tested (n = 5000) on simulated brain spectra with wide ranges of SNR (6.90–20.74) and linewidth (10–20 Hz). The CNN was further tested on in vivo spectra (n = 40) from five healthy volunteers with substantially different SNR, and the results were compared with those from the LCModel analysis. A Student t test was performed for the comparison. Results: Using the proposed method the mean-absolute-percent-errors (MAPEs) in the estimated metabolite concentrations were 12.49% ± 4.35% for aspartate, creatine (Cr), γ-aminobutyric acid (GABA), glucose, glutamine, glutamate, glutathione (GSH), myo-Inositol (mI), N-acetylaspartate, phosphocreatine (PCr), phosphorylethanolamine, and taurine over the whole simulated spectra in the test set. The metabolite concentrations estimated from in vivo spectra were close to the reported ranges for the proposed method and the LCModel analysis except mI, GSH, and especially Cr/PCr for the LCModel analysis, and phosphorylcholine to glycerophosphorylcholine ratio (PC/GPC) for both methods. The metabolite concentrations estimated across the in vivo spectra with different SNR were less variable with the proposed method (~10% or less) than with the LCModel analysis. Conclusion: The robust performance of the proposed method against low SNR may allow a subminute 1 H-MRS of human brain, which is an important technical development for clinical studies.

KW - brain

KW - convolutional neural network

KW - deep learning

KW - metabolite quantification

KW - proton magnetic resonance spectroscopy

UR - http://www.scopus.com/inward/record.url?scp=85062802474&partnerID=8YFLogxK

U2 - 10.1002/mrm.27727

DO - 10.1002/mrm.27727

M3 - Article

VL - 82

SP - 33

EP - 48

JO - Magnetic Resonance in Medicine

JF - Magnetic Resonance in Medicine

SN - 0740-3194

IS - 1

ER -