Evaluation of a deep learning approach for the segmentation of brain tissues and white matter hyperintensities of presumed vascular origin in MRI

P. Moeskops, J. de Bresser, H.J. Kuijf, A.M. Mendrik, G.J. Biessels, J.P.W. Pluim, I. Išgum

Research output: Contribution to journalArticleAcademicpeer-review

15 Citations (Scopus)
104 Downloads (Pure)

Abstract

Automatic segmentation of brain tissues and white matter hyperintensities of presumed vascular origin (WMH) in MRI of older patients is widely described in the literature. Although brain abnormalities and motion artefacts are common in this age group, most segmentation methods are not evaluated in a setting that includes these items. In the present study, our tissue segmentation method for brain MRI was extended and evaluated for additional WMH segmentation. Furthermore, our method was evaluated in two large cohorts with a realistic variation in brain abnormalities and motion artefacts. The method uses a multi-scale convolutional neural network with a T1-weighted image, a T2-weighted fluid attenuated inversion recovery (FLAIR) image and a T1-weighted inversion recovery (IR) image as input. The method automatically segments white matter (WM), cortical grey matter (cGM), basal ganglia and thalami (BGT), cerebellum (CB), brain stem (BS), lateral ventricular cerebrospinal fluid (lvCSF), peripheral cerebrospinal fluid (pCSF), and WMH. Our method was evaluated quantitatively with images publicly available from the MRBrainS13 challenge (n = 20), quantitatively and qualitatively in relatively healthy older subjects (n = 96), and qualitatively in patients from a memory clinic (n = 110). The method can accurately segment WMH (Overall Dice coefficient in the MRBrainS13 data of 0.67) without compromising performance for tissue segmentations (Overall Dice coefficients in the MRBrainS13 data of 0.87 for WM, 0.85 for cGM, 0.82 for BGT, 0.93 for CB, 0.92 for BS, 0.93 for lvCSF, 0.76 for pCSF). Furthermore, the automatic WMH volumes showed a high correlation with manual WMH volumes (Spearman's ρ = 0.83 for relatively healthy older subjects). In both cohorts, our method produced reliable segmentations (as determined by a human observer) in most images (relatively healthy/memory clinic: tissues 88%/77% reliable, WMH 85%/84% reliable) despite various degrees of brain abnormalities and motion artefacts. In conclusion, this study shows that a convolutional neural network-based segmentation method can accurately segment brain tissues and WMH in MR images of older patients with varying degrees of brain abnormalities and motion artefacts.

Original languageEnglish
Pages (from-to)251-262
Number of pages12
JournalNeuroImage: Clinical
Volume17
DOIs
Publication statusPublished - 2018

Fingerprint

Blood Vessels
Learning
Brain
Artifacts
Cerebrospinal Fluid
Basal Ganglia
Thalamus
Cerebellum
Brain Stem
Healthy Volunteers
White Matter
Age Groups

Keywords

  • Brain atrophy
  • Brain MRI
  • Convolutional neural networks
  • Deep learning
  • Motion artefacts
  • Segmentation
  • White matter hyperintensities

Cite this

@article{137af335e7f641d38e284a1ad64e2bc4,
title = "Evaluation of a deep learning approach for the segmentation of brain tissues and white matter hyperintensities of presumed vascular origin in MRI",
abstract = "Automatic segmentation of brain tissues and white matter hyperintensities of presumed vascular origin (WMH) in MRI of older patients is widely described in the literature. Although brain abnormalities and motion artefacts are common in this age group, most segmentation methods are not evaluated in a setting that includes these items. In the present study, our tissue segmentation method for brain MRI was extended and evaluated for additional WMH segmentation. Furthermore, our method was evaluated in two large cohorts with a realistic variation in brain abnormalities and motion artefacts. The method uses a multi-scale convolutional neural network with a T1-weighted image, a T2-weighted fluid attenuated inversion recovery (FLAIR) image and a T1-weighted inversion recovery (IR) image as input. The method automatically segments white matter (WM), cortical grey matter (cGM), basal ganglia and thalami (BGT), cerebellum (CB), brain stem (BS), lateral ventricular cerebrospinal fluid (lvCSF), peripheral cerebrospinal fluid (pCSF), and WMH. Our method was evaluated quantitatively with images publicly available from the MRBrainS13 challenge (n = 20), quantitatively and qualitatively in relatively healthy older subjects (n = 96), and qualitatively in patients from a memory clinic (n = 110). The method can accurately segment WMH (Overall Dice coefficient in the MRBrainS13 data of 0.67) without compromising performance for tissue segmentations (Overall Dice coefficients in the MRBrainS13 data of 0.87 for WM, 0.85 for cGM, 0.82 for BGT, 0.93 for CB, 0.92 for BS, 0.93 for lvCSF, 0.76 for pCSF). Furthermore, the automatic WMH volumes showed a high correlation with manual WMH volumes (Spearman's ρ = 0.83 for relatively healthy older subjects). In both cohorts, our method produced reliable segmentations (as determined by a human observer) in most images (relatively healthy/memory clinic: tissues 88{\%}/77{\%} reliable, WMH 85{\%}/84{\%} reliable) despite various degrees of brain abnormalities and motion artefacts. In conclusion, this study shows that a convolutional neural network-based segmentation method can accurately segment brain tissues and WMH in MR images of older patients with varying degrees of brain abnormalities and motion artefacts.",
keywords = "Brain atrophy, Brain MRI, Convolutional neural networks, Deep learning, Motion artefacts, Segmentation, White matter hyperintensities",
author = "P. Moeskops and {de Bresser}, J. and H.J. Kuijf and A.M. Mendrik and G.J. Biessels and J.P.W. Pluim and I. Išgum",
year = "2018",
doi = "10.1016/j.nicl.2017.10.007",
language = "English",
volume = "17",
pages = "251--262",
journal = "NeuroImage: Clinical",
issn = "2213-1582",
publisher = "Elsevier",

}

Evaluation of a deep learning approach for the segmentation of brain tissues and white matter hyperintensities of presumed vascular origin in MRI. / Moeskops, P.; de Bresser, J.; Kuijf, H.J.; Mendrik, A.M.; Biessels, G.J.; Pluim, J.P.W.; Išgum, I.

In: NeuroImage: Clinical, Vol. 17, 2018, p. 251-262.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Evaluation of a deep learning approach for the segmentation of brain tissues and white matter hyperintensities of presumed vascular origin in MRI

AU - Moeskops, P.

AU - de Bresser, J.

AU - Kuijf, H.J.

AU - Mendrik, A.M.

AU - Biessels, G.J.

AU - Pluim, J.P.W.

AU - Išgum, I.

PY - 2018

Y1 - 2018

N2 - Automatic segmentation of brain tissues and white matter hyperintensities of presumed vascular origin (WMH) in MRI of older patients is widely described in the literature. Although brain abnormalities and motion artefacts are common in this age group, most segmentation methods are not evaluated in a setting that includes these items. In the present study, our tissue segmentation method for brain MRI was extended and evaluated for additional WMH segmentation. Furthermore, our method was evaluated in two large cohorts with a realistic variation in brain abnormalities and motion artefacts. The method uses a multi-scale convolutional neural network with a T1-weighted image, a T2-weighted fluid attenuated inversion recovery (FLAIR) image and a T1-weighted inversion recovery (IR) image as input. The method automatically segments white matter (WM), cortical grey matter (cGM), basal ganglia and thalami (BGT), cerebellum (CB), brain stem (BS), lateral ventricular cerebrospinal fluid (lvCSF), peripheral cerebrospinal fluid (pCSF), and WMH. Our method was evaluated quantitatively with images publicly available from the MRBrainS13 challenge (n = 20), quantitatively and qualitatively in relatively healthy older subjects (n = 96), and qualitatively in patients from a memory clinic (n = 110). The method can accurately segment WMH (Overall Dice coefficient in the MRBrainS13 data of 0.67) without compromising performance for tissue segmentations (Overall Dice coefficients in the MRBrainS13 data of 0.87 for WM, 0.85 for cGM, 0.82 for BGT, 0.93 for CB, 0.92 for BS, 0.93 for lvCSF, 0.76 for pCSF). Furthermore, the automatic WMH volumes showed a high correlation with manual WMH volumes (Spearman's ρ = 0.83 for relatively healthy older subjects). In both cohorts, our method produced reliable segmentations (as determined by a human observer) in most images (relatively healthy/memory clinic: tissues 88%/77% reliable, WMH 85%/84% reliable) despite various degrees of brain abnormalities and motion artefacts. In conclusion, this study shows that a convolutional neural network-based segmentation method can accurately segment brain tissues and WMH in MR images of older patients with varying degrees of brain abnormalities and motion artefacts.

AB - Automatic segmentation of brain tissues and white matter hyperintensities of presumed vascular origin (WMH) in MRI of older patients is widely described in the literature. Although brain abnormalities and motion artefacts are common in this age group, most segmentation methods are not evaluated in a setting that includes these items. In the present study, our tissue segmentation method for brain MRI was extended and evaluated for additional WMH segmentation. Furthermore, our method was evaluated in two large cohorts with a realistic variation in brain abnormalities and motion artefacts. The method uses a multi-scale convolutional neural network with a T1-weighted image, a T2-weighted fluid attenuated inversion recovery (FLAIR) image and a T1-weighted inversion recovery (IR) image as input. The method automatically segments white matter (WM), cortical grey matter (cGM), basal ganglia and thalami (BGT), cerebellum (CB), brain stem (BS), lateral ventricular cerebrospinal fluid (lvCSF), peripheral cerebrospinal fluid (pCSF), and WMH. Our method was evaluated quantitatively with images publicly available from the MRBrainS13 challenge (n = 20), quantitatively and qualitatively in relatively healthy older subjects (n = 96), and qualitatively in patients from a memory clinic (n = 110). The method can accurately segment WMH (Overall Dice coefficient in the MRBrainS13 data of 0.67) without compromising performance for tissue segmentations (Overall Dice coefficients in the MRBrainS13 data of 0.87 for WM, 0.85 for cGM, 0.82 for BGT, 0.93 for CB, 0.92 for BS, 0.93 for lvCSF, 0.76 for pCSF). Furthermore, the automatic WMH volumes showed a high correlation with manual WMH volumes (Spearman's ρ = 0.83 for relatively healthy older subjects). In both cohorts, our method produced reliable segmentations (as determined by a human observer) in most images (relatively healthy/memory clinic: tissues 88%/77% reliable, WMH 85%/84% reliable) despite various degrees of brain abnormalities and motion artefacts. In conclusion, this study shows that a convolutional neural network-based segmentation method can accurately segment brain tissues and WMH in MR images of older patients with varying degrees of brain abnormalities and motion artefacts.

KW - Brain atrophy

KW - Brain MRI

KW - Convolutional neural networks

KW - Deep learning

KW - Motion artefacts

KW - Segmentation

KW - White matter hyperintensities

UR - http://www.scopus.com/inward/record.url?scp=85032296672&partnerID=8YFLogxK

U2 - 10.1016/j.nicl.2017.10.007

DO - 10.1016/j.nicl.2017.10.007

M3 - Article

C2 - 29159042

VL - 17

SP - 251

EP - 262

JO - NeuroImage: Clinical

JF - NeuroImage: Clinical

SN - 2213-1582

ER -