Assessing the strength of directed influences among neural signals

An approach to noisy data

Linda Sommerlade, Marco Thiel, Malenka Mader, Wolfgang Mader, Jens Timmer, Bettina Platt, Bjoern Schelter*

*Corresponding author for this work

Research output: Contribution to journalArticle

10 Citations (Scopus)
3 Downloads (Pure)

Abstract

Background: Measurements in the neurosciences are afflicted with observational noise. Granger-causality inference typically does not take this effect into account. We demonstrate that this leads to false positives conclusions and spurious causalities.

New method: State space modelling provides a convenient framework to obtain reliable estimates for Granger-causality. Despite its previous application in several studies, the analytical derivation of the statistics for parameter estimation in the state space model was missing. This prevented a rigorous evaluation of the results.

Results: In this manuscript we derive the statistics for parameter estimation in the state space model. We demonstrate in an extensive simulation study that our novel approach outperforms standard approaches and avoids false positive conclusions about Granger-causality.

Comparison with existing methods: In comparison with the naive application of Granger-causality inference, we demonstrate the superiority of our novel approach. The wide-spread applicability of our procedure provides a statistical framework for future studies. The application to mice electroencephalogram data demonstrates the immediate applicability of our approach.

Conclusions: The analytical derivation of the statistics presented in this manuscript enables a rigorous evaluation of the results of Granger causal network inference. It is noteworthy that the statistics can be readily applied to various measures for Granger causality and other approaches that are based on vector autoregressive models. (C) 2014 Elsevier B.V. All rights reserved.

Original languageEnglish
Pages (from-to)47-64
Number of pages18
JournalJournal of Neuroscience Methods
Volume239
Early online date23 Sep 2014
DOIs
Publication statusPublished - 15 Jan 2015

Keywords

  • Granger-causality
  • Observational noise
  • Statistics
  • Expectation-maximisation algorithm
  • Kalman filter
  • Incomplete data likelihood
  • Analytical covariance matrix
  • Multivariate time-series
  • Granger causality
  • Maximum-likelihood
  • Linear-dependence
  • Information-flow
  • Coherence
  • EEG
  • Feedback
  • Interval
  • Algorithm

Cite this

Assessing the strength of directed influences among neural signals : An approach to noisy data. / Sommerlade, Linda; Thiel, Marco; Mader, Malenka; Mader, Wolfgang; Timmer, Jens; Platt, Bettina; Schelter, Bjoern.

In: Journal of Neuroscience Methods, Vol. 239, 15.01.2015, p. 47-64.

Research output: Contribution to journalArticle

@article{262962e6205d492b92fe0f1c3a035f28,
title = "Assessing the strength of directed influences among neural signals: An approach to noisy data",
abstract = "Background: Measurements in the neurosciences are afflicted with observational noise. Granger-causality inference typically does not take this effect into account. We demonstrate that this leads to false positives conclusions and spurious causalities.New method: State space modelling provides a convenient framework to obtain reliable estimates for Granger-causality. Despite its previous application in several studies, the analytical derivation of the statistics for parameter estimation in the state space model was missing. This prevented a rigorous evaluation of the results.Results: In this manuscript we derive the statistics for parameter estimation in the state space model. We demonstrate in an extensive simulation study that our novel approach outperforms standard approaches and avoids false positive conclusions about Granger-causality.Comparison with existing methods: In comparison with the naive application of Granger-causality inference, we demonstrate the superiority of our novel approach. The wide-spread applicability of our procedure provides a statistical framework for future studies. The application to mice electroencephalogram data demonstrates the immediate applicability of our approach.Conclusions: The analytical derivation of the statistics presented in this manuscript enables a rigorous evaluation of the results of Granger causal network inference. It is noteworthy that the statistics can be readily applied to various measures for Granger causality and other approaches that are based on vector autoregressive models. (C) 2014 Elsevier B.V. All rights reserved.",
keywords = "Granger-causality, Observational noise, Statistics, Expectation-maximisation algorithm, Kalman filter, Incomplete data likelihood, Analytical covariance matrix, Multivariate time-series, Granger causality, Maximum-likelihood, Linear-dependence, Information-flow, Coherence, EEG, Feedback, Interval, Algorithm",
author = "Linda Sommerlade and Marco Thiel and Malenka Mader and Wolfgang Mader and Jens Timmer and Bettina Platt and Bjoern Schelter",
note = "Acknowledgements This work was supported by the German Science Foundation (Ti315/4-2), the German Federal Ministry of Education and Research (BMBF grant 01GQ0420), and the Excellence Initiative of the German Federal and State Governments. B.S. is indebted to the Kosterlitz Centre for the financial support of this research project.",
year = "2015",
month = "1",
day = "15",
doi = "10.1016/j.jneumeth.2014.09.007",
language = "English",
volume = "239",
pages = "47--64",
journal = "Journal of Neuroscience Methods",
issn = "0165-0270",
publisher = "Elsevier",

}

TY - JOUR

T1 - Assessing the strength of directed influences among neural signals

T2 - An approach to noisy data

AU - Sommerlade, Linda

AU - Thiel, Marco

AU - Mader, Malenka

AU - Mader, Wolfgang

AU - Timmer, Jens

AU - Platt, Bettina

AU - Schelter, Bjoern

N1 - Acknowledgements This work was supported by the German Science Foundation (Ti315/4-2), the German Federal Ministry of Education and Research (BMBF grant 01GQ0420), and the Excellence Initiative of the German Federal and State Governments. B.S. is indebted to the Kosterlitz Centre for the financial support of this research project.

PY - 2015/1/15

Y1 - 2015/1/15

N2 - Background: Measurements in the neurosciences are afflicted with observational noise. Granger-causality inference typically does not take this effect into account. We demonstrate that this leads to false positives conclusions and spurious causalities.New method: State space modelling provides a convenient framework to obtain reliable estimates for Granger-causality. Despite its previous application in several studies, the analytical derivation of the statistics for parameter estimation in the state space model was missing. This prevented a rigorous evaluation of the results.Results: In this manuscript we derive the statistics for parameter estimation in the state space model. We demonstrate in an extensive simulation study that our novel approach outperforms standard approaches and avoids false positive conclusions about Granger-causality.Comparison with existing methods: In comparison with the naive application of Granger-causality inference, we demonstrate the superiority of our novel approach. The wide-spread applicability of our procedure provides a statistical framework for future studies. The application to mice electroencephalogram data demonstrates the immediate applicability of our approach.Conclusions: The analytical derivation of the statistics presented in this manuscript enables a rigorous evaluation of the results of Granger causal network inference. It is noteworthy that the statistics can be readily applied to various measures for Granger causality and other approaches that are based on vector autoregressive models. (C) 2014 Elsevier B.V. All rights reserved.

AB - Background: Measurements in the neurosciences are afflicted with observational noise. Granger-causality inference typically does not take this effect into account. We demonstrate that this leads to false positives conclusions and spurious causalities.New method: State space modelling provides a convenient framework to obtain reliable estimates for Granger-causality. Despite its previous application in several studies, the analytical derivation of the statistics for parameter estimation in the state space model was missing. This prevented a rigorous evaluation of the results.Results: In this manuscript we derive the statistics for parameter estimation in the state space model. We demonstrate in an extensive simulation study that our novel approach outperforms standard approaches and avoids false positive conclusions about Granger-causality.Comparison with existing methods: In comparison with the naive application of Granger-causality inference, we demonstrate the superiority of our novel approach. The wide-spread applicability of our procedure provides a statistical framework for future studies. The application to mice electroencephalogram data demonstrates the immediate applicability of our approach.Conclusions: The analytical derivation of the statistics presented in this manuscript enables a rigorous evaluation of the results of Granger causal network inference. It is noteworthy that the statistics can be readily applied to various measures for Granger causality and other approaches that are based on vector autoregressive models. (C) 2014 Elsevier B.V. All rights reserved.

KW - Granger-causality

KW - Observational noise

KW - Statistics

KW - Expectation-maximisation algorithm

KW - Kalman filter

KW - Incomplete data likelihood

KW - Analytical covariance matrix

KW - Multivariate time-series

KW - Granger causality

KW - Maximum-likelihood

KW - Linear-dependence

KW - Information-flow

KW - Coherence

KW - EEG

KW - Feedback

KW - Interval

KW - Algorithm

U2 - 10.1016/j.jneumeth.2014.09.007

DO - 10.1016/j.jneumeth.2014.09.007

M3 - Article

VL - 239

SP - 47

EP - 64

JO - Journal of Neuroscience Methods

JF - Journal of Neuroscience Methods

SN - 0165-0270

ER -