A nonlinear causality measure in the frequency domain: Nonlinear partial directed coherence with applications to EEG

Fei He, Stephen A. Billings, Hua Liang Wei, Ptolemaios G. Sarrigiannis

Research output: Contribution to journalArticle

14 Citations (Scopus)
12 Downloads (Pure)

Abstract

Background: Frequency domain Granger causality measures have been proposed and widely applied in analyzing rhythmic neurophysiological and biomedical signals. Almost all these measures are based on linear time domain regression models, and therefore can only detect linear causal effects in the frequency domain. New method: A frequency domain causality measure, the partial directed coherence, is explicitly linked with the frequency response function concept of linear systems. By modeling the nonlinear relationships between time series using nonlinear models and employing corresponding frequency-domain analysis techniques (i.e. generalized frequency response functions), a new nonlinear partial directed coherence method is derived. Results: The advantages of the new method are illustrated via a numerical example of a nonlinear physical system and an application to electroencephalogram signals from a patient with childhood absence epilepsy. Comparison with existing methods: The new method detects both linear and nonlinear casual effects between bivariate signals in the frequency domain, while the existing measures can only detect linear effects. Conclusions: The proposed new method has important advantages over the classical linear measures, because detecting nonlinear dependencies has become more and more important in characterizing functional couplings in neuronal and biological systems.

Original languageEnglish
Pages (from-to)71-80
Number of pages10
JournalJournal of Neuroscience Methods
Volume225
Early online date25 Jan 2014
DOIs
Publication statusPublished - 30 Mar 2014
Externally publishedYes

Fingerprint

Causality
Electroencephalography
Absence Epilepsy
Nonlinear Dynamics

Keywords

  • Coherence
  • Epilepsy
  • Granger causality
  • Nonlinear systems
  • Spectral analysis

ASJC Scopus subject areas

  • Neuroscience(all)

Cite this

A nonlinear causality measure in the frequency domain : Nonlinear partial directed coherence with applications to EEG. / He, Fei; Billings, Stephen A.; Wei, Hua Liang; Sarrigiannis, Ptolemaios G.

In: Journal of Neuroscience Methods, Vol. 225, 30.03.2014, p. 71-80.

Research output: Contribution to journalArticle

He, Fei ; Billings, Stephen A. ; Wei, Hua Liang ; Sarrigiannis, Ptolemaios G. / A nonlinear causality measure in the frequency domain : Nonlinear partial directed coherence with applications to EEG. In: Journal of Neuroscience Methods. 2014 ; Vol. 225. pp. 71-80.
@article{a39ce8a6af5f4b97b0322a50a61b9d06,
title = "A nonlinear causality measure in the frequency domain: Nonlinear partial directed coherence with applications to EEG",
abstract = "Background: Frequency domain Granger causality measures have been proposed and widely applied in analyzing rhythmic neurophysiological and biomedical signals. Almost all these measures are based on linear time domain regression models, and therefore can only detect linear causal effects in the frequency domain. New method: A frequency domain causality measure, the partial directed coherence, is explicitly linked with the frequency response function concept of linear systems. By modeling the nonlinear relationships between time series using nonlinear models and employing corresponding frequency-domain analysis techniques (i.e. generalized frequency response functions), a new nonlinear partial directed coherence method is derived. Results: The advantages of the new method are illustrated via a numerical example of a nonlinear physical system and an application to electroencephalogram signals from a patient with childhood absence epilepsy. Comparison with existing methods: The new method detects both linear and nonlinear casual effects between bivariate signals in the frequency domain, while the existing measures can only detect linear effects. Conclusions: The proposed new method has important advantages over the classical linear measures, because detecting nonlinear dependencies has become more and more important in characterizing functional couplings in neuronal and biological systems.",
keywords = "Coherence, Epilepsy, Granger causality, Nonlinear systems, Spectral analysis",
author = "Fei He and Billings, {Stephen A.} and Wei, {Hua Liang} and Sarrigiannis, {Ptolemaios G.}",
year = "2014",
month = "3",
day = "30",
doi = "10.1016/j.jneumeth.2014.01.013",
language = "English",
volume = "225",
pages = "71--80",
journal = "Journal of Neuroscience Methods",
issn = "0165-0270",
publisher = "Elsevier",

}

TY - JOUR

T1 - A nonlinear causality measure in the frequency domain

T2 - Nonlinear partial directed coherence with applications to EEG

AU - He, Fei

AU - Billings, Stephen A.

AU - Wei, Hua Liang

AU - Sarrigiannis, Ptolemaios G.

PY - 2014/3/30

Y1 - 2014/3/30

N2 - Background: Frequency domain Granger causality measures have been proposed and widely applied in analyzing rhythmic neurophysiological and biomedical signals. Almost all these measures are based on linear time domain regression models, and therefore can only detect linear causal effects in the frequency domain. New method: A frequency domain causality measure, the partial directed coherence, is explicitly linked with the frequency response function concept of linear systems. By modeling the nonlinear relationships between time series using nonlinear models and employing corresponding frequency-domain analysis techniques (i.e. generalized frequency response functions), a new nonlinear partial directed coherence method is derived. Results: The advantages of the new method are illustrated via a numerical example of a nonlinear physical system and an application to electroencephalogram signals from a patient with childhood absence epilepsy. Comparison with existing methods: The new method detects both linear and nonlinear casual effects between bivariate signals in the frequency domain, while the existing measures can only detect linear effects. Conclusions: The proposed new method has important advantages over the classical linear measures, because detecting nonlinear dependencies has become more and more important in characterizing functional couplings in neuronal and biological systems.

AB - Background: Frequency domain Granger causality measures have been proposed and widely applied in analyzing rhythmic neurophysiological and biomedical signals. Almost all these measures are based on linear time domain regression models, and therefore can only detect linear causal effects in the frequency domain. New method: A frequency domain causality measure, the partial directed coherence, is explicitly linked with the frequency response function concept of linear systems. By modeling the nonlinear relationships between time series using nonlinear models and employing corresponding frequency-domain analysis techniques (i.e. generalized frequency response functions), a new nonlinear partial directed coherence method is derived. Results: The advantages of the new method are illustrated via a numerical example of a nonlinear physical system and an application to electroencephalogram signals from a patient with childhood absence epilepsy. Comparison with existing methods: The new method detects both linear and nonlinear casual effects between bivariate signals in the frequency domain, while the existing measures can only detect linear effects. Conclusions: The proposed new method has important advantages over the classical linear measures, because detecting nonlinear dependencies has become more and more important in characterizing functional couplings in neuronal and biological systems.

KW - Coherence

KW - Epilepsy

KW - Granger causality

KW - Nonlinear systems

KW - Spectral analysis

UR - http://www.scopus.com/inward/record.url?scp=84893677832&partnerID=8YFLogxK

U2 - 10.1016/j.jneumeth.2014.01.013

DO - 10.1016/j.jneumeth.2014.01.013

M3 - Article

VL - 225

SP - 71

EP - 80

JO - Journal of Neuroscience Methods

JF - Journal of Neuroscience Methods

SN - 0165-0270

ER -