Multiple functional networks modeling for autism spectrum disorder diagnosis

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Despite countless studies on autism spectrum disorder (ASD), diagnosis relies on specific behavioral criteria and neuroimaging biomarkers for the disorder are still relatively scarce and irrelevant for diagnostic workup. Many researchers have focused on functional networks of brain activities using resting-state functional magnetic resonance imaging (rsfMRI) to diagnose brain diseases, including ASD. Although some existing methods are able to reveal the abnormalities in functional networks, they are either highly dependent on prior assumptions for modeling these networks or do not focus on latent functional connectivities (FCs) by considering discriminative relations among FCs in a nonlinear way. In this article, we propose a novel framework to model multiple networks of rsfMRI with data-driven approaches. Specifically, we construct large-scale functional networks with hierarchical clustering and find discriminative connectivity patterns between ASD and normal controls (NC). We then learn features and classifiers for each cluster through discriminative restricted Boltzmann machines (DRBMs). In the testing phase, each DRBM determines whether a test sample is ASD or NC, based on which we make a final decision with a majority voting strategy. We assess the diagnostic performance of the proposed method using public datasets and describe the effectiveness of our method by comparing it to competing methods. We also rigorously analyze FCs learned by DRBMs on each cluster and discover dominant FCs that play a major role in discriminating between ASD and NC. Hum Brain Mapp 38:5804–5821, 2017.

Original languageEnglish
Pages (from-to)5804-5821
Number of pages18
JournalHuman Brain Mapping
Volume38
Issue number11
DOIs
Publication statusPublished - 2017 Nov 1

Fingerprint

Magnetic Resonance Imaging
Brain
Brain Diseases
Politics
Neuroimaging
Cluster Analysis
Biomarkers
Research Personnel
Autism Spectrum Disorder
Datasets

Keywords

  • autism spectrum disorder
  • discriminative restricted Boltzmann machine
  • functional magnetic resonance imaging
  • functional network analysis
  • hierarchical clustering
  • multiple clusters

ASJC Scopus subject areas

  • Anatomy
  • Radiological and Ultrasound Technology
  • Radiology Nuclear Medicine and imaging
  • Neurology
  • Clinical Neurology

Cite this

Multiple functional networks modeling for autism spectrum disorder diagnosis. / Kam, Tae Eui; Suk, Heung-Il; Lee, Seong Whan.

In: Human Brain Mapping, Vol. 38, No. 11, 01.11.2017, p. 5804-5821.

Research output: Contribution to journalArticle

@article{814b6f6c36064d159bfe8a7cb55a42cb,
title = "Multiple functional networks modeling for autism spectrum disorder diagnosis",
abstract = "Despite countless studies on autism spectrum disorder (ASD), diagnosis relies on specific behavioral criteria and neuroimaging biomarkers for the disorder are still relatively scarce and irrelevant for diagnostic workup. Many researchers have focused on functional networks of brain activities using resting-state functional magnetic resonance imaging (rsfMRI) to diagnose brain diseases, including ASD. Although some existing methods are able to reveal the abnormalities in functional networks, they are either highly dependent on prior assumptions for modeling these networks or do not focus on latent functional connectivities (FCs) by considering discriminative relations among FCs in a nonlinear way. In this article, we propose a novel framework to model multiple networks of rsfMRI with data-driven approaches. Specifically, we construct large-scale functional networks with hierarchical clustering and find discriminative connectivity patterns between ASD and normal controls (NC). We then learn features and classifiers for each cluster through discriminative restricted Boltzmann machines (DRBMs). In the testing phase, each DRBM determines whether a test sample is ASD or NC, based on which we make a final decision with a majority voting strategy. We assess the diagnostic performance of the proposed method using public datasets and describe the effectiveness of our method by comparing it to competing methods. We also rigorously analyze FCs learned by DRBMs on each cluster and discover dominant FCs that play a major role in discriminating between ASD and NC. Hum Brain Mapp 38:5804–5821, 2017.",
keywords = "autism spectrum disorder, discriminative restricted Boltzmann machine, functional magnetic resonance imaging, functional network analysis, hierarchical clustering, multiple clusters",
author = "Kam, {Tae Eui} and Heung-Il Suk and Lee, {Seong Whan}",
year = "2017",
month = "11",
day = "1",
doi = "10.1002/hbm.23769",
language = "English",
volume = "38",
pages = "5804--5821",
journal = "Human Brain Mapping",
issn = "1065-9471",
publisher = "Wiley-Liss Inc.",
number = "11",

}

TY - JOUR

T1 - Multiple functional networks modeling for autism spectrum disorder diagnosis

AU - Kam, Tae Eui

AU - Suk, Heung-Il

AU - Lee, Seong Whan

PY - 2017/11/1

Y1 - 2017/11/1

N2 - Despite countless studies on autism spectrum disorder (ASD), diagnosis relies on specific behavioral criteria and neuroimaging biomarkers for the disorder are still relatively scarce and irrelevant for diagnostic workup. Many researchers have focused on functional networks of brain activities using resting-state functional magnetic resonance imaging (rsfMRI) to diagnose brain diseases, including ASD. Although some existing methods are able to reveal the abnormalities in functional networks, they are either highly dependent on prior assumptions for modeling these networks or do not focus on latent functional connectivities (FCs) by considering discriminative relations among FCs in a nonlinear way. In this article, we propose a novel framework to model multiple networks of rsfMRI with data-driven approaches. Specifically, we construct large-scale functional networks with hierarchical clustering and find discriminative connectivity patterns between ASD and normal controls (NC). We then learn features and classifiers for each cluster through discriminative restricted Boltzmann machines (DRBMs). In the testing phase, each DRBM determines whether a test sample is ASD or NC, based on which we make a final decision with a majority voting strategy. We assess the diagnostic performance of the proposed method using public datasets and describe the effectiveness of our method by comparing it to competing methods. We also rigorously analyze FCs learned by DRBMs on each cluster and discover dominant FCs that play a major role in discriminating between ASD and NC. Hum Brain Mapp 38:5804–5821, 2017.

AB - Despite countless studies on autism spectrum disorder (ASD), diagnosis relies on specific behavioral criteria and neuroimaging biomarkers for the disorder are still relatively scarce and irrelevant for diagnostic workup. Many researchers have focused on functional networks of brain activities using resting-state functional magnetic resonance imaging (rsfMRI) to diagnose brain diseases, including ASD. Although some existing methods are able to reveal the abnormalities in functional networks, they are either highly dependent on prior assumptions for modeling these networks or do not focus on latent functional connectivities (FCs) by considering discriminative relations among FCs in a nonlinear way. In this article, we propose a novel framework to model multiple networks of rsfMRI with data-driven approaches. Specifically, we construct large-scale functional networks with hierarchical clustering and find discriminative connectivity patterns between ASD and normal controls (NC). We then learn features and classifiers for each cluster through discriminative restricted Boltzmann machines (DRBMs). In the testing phase, each DRBM determines whether a test sample is ASD or NC, based on which we make a final decision with a majority voting strategy. We assess the diagnostic performance of the proposed method using public datasets and describe the effectiveness of our method by comparing it to competing methods. We also rigorously analyze FCs learned by DRBMs on each cluster and discover dominant FCs that play a major role in discriminating between ASD and NC. Hum Brain Mapp 38:5804–5821, 2017.

KW - autism spectrum disorder

KW - discriminative restricted Boltzmann machine

KW - functional magnetic resonance imaging

KW - functional network analysis

KW - hierarchical clustering

KW - multiple clusters

UR - http://www.scopus.com/inward/record.url?scp=85038014394&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85038014394&partnerID=8YFLogxK

U2 - 10.1002/hbm.23769

DO - 10.1002/hbm.23769

M3 - Article

C2 - 28845892

AN - SCOPUS:85038014394

VL - 38

SP - 5804

EP - 5821

JO - Human Brain Mapping

JF - Human Brain Mapping

SN - 1065-9471

IS - 11

ER -