Test–retest reliability of spatial patterns from resting-state functional MRI using the restricted Boltzmann machine and hierarchically organized spatial patterns from the deep belief network

Hyun Chul Kim, Hojin Jang, Jong Hwan Lee

Research output: Contribution to journalArticle

Abstract

Background: Restricted Boltzmann machines (RBMs), including greedy layer-wise trained RBMs as part of a deep belief network (DBN), have the ability to identify spatial patterns (SPs; functional networks) in resting-state fMRI (rfMRI) data. However, there has been little research on (1) the reproducibility and test-retest reliability of SPs derived from RBMs and on (2) hierarchical SPs derived from DBNs. Methods: We applied a weight sparsity-controlled RBM and DBN to whole-brain rfMRI data from the Human Connectome Project. We evaluated the within-session reproducibility and between-session test-retest reliability of the SPs derived from the RBM approach and compared them both with those identified using independent component analysis (ICA) and with three voxel-wise statistical measures—the Hurst exponent, entropy, and kurtosis—of the rfMRI data. We also assessed the potential hierarchy of the SPs from the DBN. Results: An increase in the sparsity level of the RBM weights enhanced the reproducibility of the SPs. The SPs deriving from a stringent weight sparsity level were predominantly found in the cortical gray matter and substantially overlapped with the SPs obtained from the Hurst exponent. A hierarchical representation was shown by constructed using the default-mode network obtained from the DBN. Comparison with existing methods: The test-retest reliability of the SPs from the RBM was superior to that of the SPs from the voxel-wise statistics. Conclusions: The SPs from the RBM were reproducible within sessions and reliable across sessions. The hierarchically organized SPs from the DBN could possibly be applied to research based on rfMRI data.

Original languageEnglish
Article number108451
JournalJournal of Neuroscience Methods
Volume330
DOIs
Publication statusPublished - 2020 Jan 15

Fingerprint

Magnetic Resonance Imaging
Reproducibility of Results
Weights and Measures
Connectome
Entropy
Research
Brain

Keywords

  • Deep belief network
  • Entropy
  • Hurst exponent
  • Independent component analysis
  • Kurtosis
  • Resting-state fMRI
  • Restricted Boltzmann machine

ASJC Scopus subject areas

  • Neuroscience(all)

Cite this

@article{f8b662914ba64c7b86c20797d0e1c529,
title = "Test–retest reliability of spatial patterns from resting-state functional MRI using the restricted Boltzmann machine and hierarchically organized spatial patterns from the deep belief network",
abstract = "Background: Restricted Boltzmann machines (RBMs), including greedy layer-wise trained RBMs as part of a deep belief network (DBN), have the ability to identify spatial patterns (SPs; functional networks) in resting-state fMRI (rfMRI) data. However, there has been little research on (1) the reproducibility and test-retest reliability of SPs derived from RBMs and on (2) hierarchical SPs derived from DBNs. Methods: We applied a weight sparsity-controlled RBM and DBN to whole-brain rfMRI data from the Human Connectome Project. We evaluated the within-session reproducibility and between-session test-retest reliability of the SPs derived from the RBM approach and compared them both with those identified using independent component analysis (ICA) and with three voxel-wise statistical measures—the Hurst exponent, entropy, and kurtosis—of the rfMRI data. We also assessed the potential hierarchy of the SPs from the DBN. Results: An increase in the sparsity level of the RBM weights enhanced the reproducibility of the SPs. The SPs deriving from a stringent weight sparsity level were predominantly found in the cortical gray matter and substantially overlapped with the SPs obtained from the Hurst exponent. A hierarchical representation was shown by constructed using the default-mode network obtained from the DBN. Comparison with existing methods: The test-retest reliability of the SPs from the RBM was superior to that of the SPs from the voxel-wise statistics. Conclusions: The SPs from the RBM were reproducible within sessions and reliable across sessions. The hierarchically organized SPs from the DBN could possibly be applied to research based on rfMRI data.",
keywords = "Deep belief network, Entropy, Hurst exponent, Independent component analysis, Kurtosis, Resting-state fMRI, Restricted Boltzmann machine",
author = "Kim, {Hyun Chul} and Hojin Jang and Lee, {Jong Hwan}",
year = "2020",
month = "1",
day = "15",
doi = "10.1016/j.jneumeth.2019.108451",
language = "English",
volume = "330",
journal = "Journal of Neuroscience Methods",
issn = "0165-0270",
publisher = "Elsevier",

}

TY - JOUR

T1 - Test–retest reliability of spatial patterns from resting-state functional MRI using the restricted Boltzmann machine and hierarchically organized spatial patterns from the deep belief network

AU - Kim, Hyun Chul

AU - Jang, Hojin

AU - Lee, Jong Hwan

PY - 2020/1/15

Y1 - 2020/1/15

N2 - Background: Restricted Boltzmann machines (RBMs), including greedy layer-wise trained RBMs as part of a deep belief network (DBN), have the ability to identify spatial patterns (SPs; functional networks) in resting-state fMRI (rfMRI) data. However, there has been little research on (1) the reproducibility and test-retest reliability of SPs derived from RBMs and on (2) hierarchical SPs derived from DBNs. Methods: We applied a weight sparsity-controlled RBM and DBN to whole-brain rfMRI data from the Human Connectome Project. We evaluated the within-session reproducibility and between-session test-retest reliability of the SPs derived from the RBM approach and compared them both with those identified using independent component analysis (ICA) and with three voxel-wise statistical measures—the Hurst exponent, entropy, and kurtosis—of the rfMRI data. We also assessed the potential hierarchy of the SPs from the DBN. Results: An increase in the sparsity level of the RBM weights enhanced the reproducibility of the SPs. The SPs deriving from a stringent weight sparsity level were predominantly found in the cortical gray matter and substantially overlapped with the SPs obtained from the Hurst exponent. A hierarchical representation was shown by constructed using the default-mode network obtained from the DBN. Comparison with existing methods: The test-retest reliability of the SPs from the RBM was superior to that of the SPs from the voxel-wise statistics. Conclusions: The SPs from the RBM were reproducible within sessions and reliable across sessions. The hierarchically organized SPs from the DBN could possibly be applied to research based on rfMRI data.

AB - Background: Restricted Boltzmann machines (RBMs), including greedy layer-wise trained RBMs as part of a deep belief network (DBN), have the ability to identify spatial patterns (SPs; functional networks) in resting-state fMRI (rfMRI) data. However, there has been little research on (1) the reproducibility and test-retest reliability of SPs derived from RBMs and on (2) hierarchical SPs derived from DBNs. Methods: We applied a weight sparsity-controlled RBM and DBN to whole-brain rfMRI data from the Human Connectome Project. We evaluated the within-session reproducibility and between-session test-retest reliability of the SPs derived from the RBM approach and compared them both with those identified using independent component analysis (ICA) and with three voxel-wise statistical measures—the Hurst exponent, entropy, and kurtosis—of the rfMRI data. We also assessed the potential hierarchy of the SPs from the DBN. Results: An increase in the sparsity level of the RBM weights enhanced the reproducibility of the SPs. The SPs deriving from a stringent weight sparsity level were predominantly found in the cortical gray matter and substantially overlapped with the SPs obtained from the Hurst exponent. A hierarchical representation was shown by constructed using the default-mode network obtained from the DBN. Comparison with existing methods: The test-retest reliability of the SPs from the RBM was superior to that of the SPs from the voxel-wise statistics. Conclusions: The SPs from the RBM were reproducible within sessions and reliable across sessions. The hierarchically organized SPs from the DBN could possibly be applied to research based on rfMRI data.

KW - Deep belief network

KW - Entropy

KW - Hurst exponent

KW - Independent component analysis

KW - Kurtosis

KW - Resting-state fMRI

KW - Restricted Boltzmann machine

UR - http://www.scopus.com/inward/record.url?scp=85074409670&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85074409670&partnerID=8YFLogxK

U2 - 10.1016/j.jneumeth.2019.108451

DO - 10.1016/j.jneumeth.2019.108451

M3 - Article

C2 - 31626847

AN - SCOPUS:85074409670

VL - 330

JO - Journal of Neuroscience Methods

JF - Journal of Neuroscience Methods

SN - 0165-0270

M1 - 108451

ER -