Extraction of Time-Varying Spatiotemporal Networks Using Parameter-Tuned Constrained IVA

Suchita Bhinge, Rami Mowakeaa, Vince D. Calhoun, Tülay Adali

Research output: Contribution to journalArticle

Abstract

Dynamic functional connectivity analysis is an effective way to capture the networks that are functionally associated and continuously changing over the scanning period. However, these methods mostly analyze the dynamic associations across the activation patterns of the spatial networks while assuming that the spatial networks are stationary. Hence, a model that allows for the variability in both domains and reduces the assumptions imposed on the data provides an effective way for extracting spatiotemporal networks. Independent vector analysis (IVA) is a joint blind source separation technique that allows for estimation of spatial and temporal features while successfully preserving variability. However, its performance is affected for higher number of datasets. Hence, we develop an effective two-stage method to extract time-varying spatial and temporal features using IVA, mitigating the problems with higher number of datasets while preserving the variability across subjects and time. The first stage is used to extract reference signals using group-independent component analysis (GICA) that are used in a parameter-tuned constrained IVA framework to estimate time-varying representations of these signals by preserving the variability through tuning the constraint parameter. This approach effectively captures variability across time from a large-scale resting-state fMRI data acquired from healthy controls and patients with schizophrenia and identifies more functionally relevant connections that are significantly different among healthy controls and patients with schizophrenia, compared with the widely used GICA method alone.

Original languageEnglish (US)
Article number8624617
Pages (from-to)1715-1725
Number of pages11
JournalIEEE transactions on medical imaging
Volume38
Issue number7
DOIs
StatePublished - Jul 1 2019
Externally publishedYes

Fingerprint

Time varying networks
Independent component analysis
Functional analysis
Blind source separation
Schizophrenia
Tuning
Chemical activation
Scanning
Joints
Magnetic Resonance Imaging

Keywords

  • Blind source separation
  • connectivity analysis
  • dimensionality reduction
  • fMRI analysis

ASJC Scopus subject areas

  • Software
  • Radiological and Ultrasound Technology
  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

Extraction of Time-Varying Spatiotemporal Networks Using Parameter-Tuned Constrained IVA. / Bhinge, Suchita; Mowakeaa, Rami; Calhoun, Vince D.; Adali, Tülay.

In: IEEE transactions on medical imaging, Vol. 38, No. 7, 8624617, 01.07.2019, p. 1715-1725.

Research output: Contribution to journalArticle

Bhinge, Suchita ; Mowakeaa, Rami ; Calhoun, Vince D. ; Adali, Tülay. / Extraction of Time-Varying Spatiotemporal Networks Using Parameter-Tuned Constrained IVA. In: IEEE transactions on medical imaging. 2019 ; Vol. 38, No. 7. pp. 1715-1725.
@article{752e608aaadc499c99345a6a2a261a00,
title = "Extraction of Time-Varying Spatiotemporal Networks Using Parameter-Tuned Constrained IVA",
abstract = "Dynamic functional connectivity analysis is an effective way to capture the networks that are functionally associated and continuously changing over the scanning period. However, these methods mostly analyze the dynamic associations across the activation patterns of the spatial networks while assuming that the spatial networks are stationary. Hence, a model that allows for the variability in both domains and reduces the assumptions imposed on the data provides an effective way for extracting spatiotemporal networks. Independent vector analysis (IVA) is a joint blind source separation technique that allows for estimation of spatial and temporal features while successfully preserving variability. However, its performance is affected for higher number of datasets. Hence, we develop an effective two-stage method to extract time-varying spatial and temporal features using IVA, mitigating the problems with higher number of datasets while preserving the variability across subjects and time. The first stage is used to extract reference signals using group-independent component analysis (GICA) that are used in a parameter-tuned constrained IVA framework to estimate time-varying representations of these signals by preserving the variability through tuning the constraint parameter. This approach effectively captures variability across time from a large-scale resting-state fMRI data acquired from healthy controls and patients with schizophrenia and identifies more functionally relevant connections that are significantly different among healthy controls and patients with schizophrenia, compared with the widely used GICA method alone.",
keywords = "Blind source separation, connectivity analysis, dimensionality reduction, fMRI analysis",
author = "Suchita Bhinge and Rami Mowakeaa and Calhoun, {Vince D.} and T{\"u}lay Adali",
year = "2019",
month = "7",
day = "1",
doi = "10.1109/TMI.2019.2893651",
language = "English (US)",
volume = "38",
pages = "1715--1725",
journal = "IEEE Transactions on Medical Imaging",
issn = "0278-0062",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "7",

}

TY - JOUR

T1 - Extraction of Time-Varying Spatiotemporal Networks Using Parameter-Tuned Constrained IVA

AU - Bhinge, Suchita

AU - Mowakeaa, Rami

AU - Calhoun, Vince D.

AU - Adali, Tülay

PY - 2019/7/1

Y1 - 2019/7/1

N2 - Dynamic functional connectivity analysis is an effective way to capture the networks that are functionally associated and continuously changing over the scanning period. However, these methods mostly analyze the dynamic associations across the activation patterns of the spatial networks while assuming that the spatial networks are stationary. Hence, a model that allows for the variability in both domains and reduces the assumptions imposed on the data provides an effective way for extracting spatiotemporal networks. Independent vector analysis (IVA) is a joint blind source separation technique that allows for estimation of spatial and temporal features while successfully preserving variability. However, its performance is affected for higher number of datasets. Hence, we develop an effective two-stage method to extract time-varying spatial and temporal features using IVA, mitigating the problems with higher number of datasets while preserving the variability across subjects and time. The first stage is used to extract reference signals using group-independent component analysis (GICA) that are used in a parameter-tuned constrained IVA framework to estimate time-varying representations of these signals by preserving the variability through tuning the constraint parameter. This approach effectively captures variability across time from a large-scale resting-state fMRI data acquired from healthy controls and patients with schizophrenia and identifies more functionally relevant connections that are significantly different among healthy controls and patients with schizophrenia, compared with the widely used GICA method alone.

AB - Dynamic functional connectivity analysis is an effective way to capture the networks that are functionally associated and continuously changing over the scanning period. However, these methods mostly analyze the dynamic associations across the activation patterns of the spatial networks while assuming that the spatial networks are stationary. Hence, a model that allows for the variability in both domains and reduces the assumptions imposed on the data provides an effective way for extracting spatiotemporal networks. Independent vector analysis (IVA) is a joint blind source separation technique that allows for estimation of spatial and temporal features while successfully preserving variability. However, its performance is affected for higher number of datasets. Hence, we develop an effective two-stage method to extract time-varying spatial and temporal features using IVA, mitigating the problems with higher number of datasets while preserving the variability across subjects and time. The first stage is used to extract reference signals using group-independent component analysis (GICA) that are used in a parameter-tuned constrained IVA framework to estimate time-varying representations of these signals by preserving the variability through tuning the constraint parameter. This approach effectively captures variability across time from a large-scale resting-state fMRI data acquired from healthy controls and patients with schizophrenia and identifies more functionally relevant connections that are significantly different among healthy controls and patients with schizophrenia, compared with the widely used GICA method alone.

KW - Blind source separation

KW - connectivity analysis

KW - dimensionality reduction

KW - fMRI analysis

UR - http://www.scopus.com/inward/record.url?scp=85068485050&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85068485050&partnerID=8YFLogxK

U2 - 10.1109/TMI.2019.2893651

DO - 10.1109/TMI.2019.2893651

M3 - Article

C2 - 30676948

AN - SCOPUS:85068485050

VL - 38

SP - 1715

EP - 1725

JO - IEEE Transactions on Medical Imaging

JF - IEEE Transactions on Medical Imaging

SN - 0278-0062

IS - 7

M1 - 8624617

ER -