Assessing uncertainty in dynamic functional connectivity

Maria Kudela, Jaroslaw Harezlak, Martin A. Lindquist

Research output: Research - peer-reviewArticle

  • 3 Citations

Abstract

Functional connectivity (FC) – the study of the statistical association between time series from anatomically distinct regions (Friston, 1994, 2011) – has become one of the primary areas of research in the field surrounding resting state functional magnetic resonance imaging (rs-fMRI). Although for many years researchers have implicitly assumed that FC was stationary across time in rs-fMRI, it has recently become increasingly clear that this is not the case and the ability to assess dynamic changes in FC is critical for better understanding of the inner workings of the human brain (Hutchison et al., 2013; Chang and Glover, 2010). Currently, the most common strategy for estimating these dynamic changes is to use the sliding-window technique. However, its greatest shortcoming is the inherent variation present in the estimate, even for null data, which is easily confused with true time-varying changes in connectivity (Lindquist et al., 2014). This can have serious consequences as even spurious fluctuations caused by noise can easily be confused with an important signal. For these reasons, assessment of uncertainty in the sliding-window correlation estimates is of critical importance. Here we propose a new approach that combines the multivariate linear process bootstrap (MLPB) method and a sliding-window technique to assess the uncertainty in a dynamic FC estimate by providing its confidence bands. Both numerical results and an application to rs-fMRI study are presented, showing the efficacy of the proposed method.

LanguageEnglish (US)
Pages165-177
Number of pages13
JournalNeuroImage
Volume149
DOIs
StatePublished - Apr 1 2017

Fingerprint

Uncertainty
Magnetic Resonance Imaging
Noise
Research Personnel
Brain
Research

Keywords

  • Dynamic confidence bands
  • Dynamic functional connectivity
  • Multivariate time series bootstrap
  • Time-varying correlation

ASJC Scopus subject areas

  • Neurology
  • Cognitive Neuroscience

Cite this

Assessing uncertainty in dynamic functional connectivity. / Kudela, Maria; Harezlak, Jaroslaw; Lindquist, Martin A.

In: NeuroImage, Vol. 149, 01.04.2017, p. 165-177.

Research output: Research - peer-reviewArticle

Kudela, Maria ; Harezlak, Jaroslaw ; Lindquist, Martin A./ Assessing uncertainty in dynamic functional connectivity. In: NeuroImage. 2017 ; Vol. 149. pp. 165-177
@article{dc789372171a4c44987cf85d75b949f1,
title = "Assessing uncertainty in dynamic functional connectivity",
abstract = "Functional connectivity (FC) – the study of the statistical association between time series from anatomically distinct regions (Friston, 1994, 2011) – has become one of the primary areas of research in the field surrounding resting state functional magnetic resonance imaging (rs-fMRI). Although for many years researchers have implicitly assumed that FC was stationary across time in rs-fMRI, it has recently become increasingly clear that this is not the case and the ability to assess dynamic changes in FC is critical for better understanding of the inner workings of the human brain (Hutchison et al., 2013; Chang and Glover, 2010). Currently, the most common strategy for estimating these dynamic changes is to use the sliding-window technique. However, its greatest shortcoming is the inherent variation present in the estimate, even for null data, which is easily confused with true time-varying changes in connectivity (Lindquist et al., 2014). This can have serious consequences as even spurious fluctuations caused by noise can easily be confused with an important signal. For these reasons, assessment of uncertainty in the sliding-window correlation estimates is of critical importance. Here we propose a new approach that combines the multivariate linear process bootstrap (MLPB) method and a sliding-window technique to assess the uncertainty in a dynamic FC estimate by providing its confidence bands. Both numerical results and an application to rs-fMRI study are presented, showing the efficacy of the proposed method.",
keywords = "Dynamic confidence bands, Dynamic functional connectivity, Multivariate time series bootstrap, Time-varying correlation",
author = "Maria Kudela and Jaroslaw Harezlak and Lindquist, {Martin A.}",
year = "2017",
month = "4",
doi = "10.1016/j.neuroimage.2017.01.056",
volume = "149",
pages = "165--177",
journal = "NeuroImage",
issn = "1053-8119",
publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - Assessing uncertainty in dynamic functional connectivity

AU - Kudela,Maria

AU - Harezlak,Jaroslaw

AU - Lindquist,Martin A.

PY - 2017/4/1

Y1 - 2017/4/1

N2 - Functional connectivity (FC) – the study of the statistical association between time series from anatomically distinct regions (Friston, 1994, 2011) – has become one of the primary areas of research in the field surrounding resting state functional magnetic resonance imaging (rs-fMRI). Although for many years researchers have implicitly assumed that FC was stationary across time in rs-fMRI, it has recently become increasingly clear that this is not the case and the ability to assess dynamic changes in FC is critical for better understanding of the inner workings of the human brain (Hutchison et al., 2013; Chang and Glover, 2010). Currently, the most common strategy for estimating these dynamic changes is to use the sliding-window technique. However, its greatest shortcoming is the inherent variation present in the estimate, even for null data, which is easily confused with true time-varying changes in connectivity (Lindquist et al., 2014). This can have serious consequences as even spurious fluctuations caused by noise can easily be confused with an important signal. For these reasons, assessment of uncertainty in the sliding-window correlation estimates is of critical importance. Here we propose a new approach that combines the multivariate linear process bootstrap (MLPB) method and a sliding-window technique to assess the uncertainty in a dynamic FC estimate by providing its confidence bands. Both numerical results and an application to rs-fMRI study are presented, showing the efficacy of the proposed method.

AB - Functional connectivity (FC) – the study of the statistical association between time series from anatomically distinct regions (Friston, 1994, 2011) – has become one of the primary areas of research in the field surrounding resting state functional magnetic resonance imaging (rs-fMRI). Although for many years researchers have implicitly assumed that FC was stationary across time in rs-fMRI, it has recently become increasingly clear that this is not the case and the ability to assess dynamic changes in FC is critical for better understanding of the inner workings of the human brain (Hutchison et al., 2013; Chang and Glover, 2010). Currently, the most common strategy for estimating these dynamic changes is to use the sliding-window technique. However, its greatest shortcoming is the inherent variation present in the estimate, even for null data, which is easily confused with true time-varying changes in connectivity (Lindquist et al., 2014). This can have serious consequences as even spurious fluctuations caused by noise can easily be confused with an important signal. For these reasons, assessment of uncertainty in the sliding-window correlation estimates is of critical importance. Here we propose a new approach that combines the multivariate linear process bootstrap (MLPB) method and a sliding-window technique to assess the uncertainty in a dynamic FC estimate by providing its confidence bands. Both numerical results and an application to rs-fMRI study are presented, showing the efficacy of the proposed method.

KW - Dynamic confidence bands

KW - Dynamic functional connectivity

KW - Multivariate time series bootstrap

KW - Time-varying correlation

UR - http://www.scopus.com/inward/record.url?scp=85012165790&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85012165790&partnerID=8YFLogxK

U2 - 10.1016/j.neuroimage.2017.01.056

DO - 10.1016/j.neuroimage.2017.01.056

M3 - Article

VL - 149

SP - 165

EP - 177

JO - NeuroImage

T2 - NeuroImage

JF - NeuroImage

SN - 1053-8119

ER -