The impact of latent confounders in directed network analysis in neuroscience

Rebecca Ramb (Corresponding Author), Michael Eichler, Alex Ing, Marco Thiel, Cornelius Weiller, Celso Grebogi, Christian Schwarzbauer, Jens Timmer, Bjoern Schelter

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

In the analysis of neuroscience data, the identification of task-related causal relationships between various areas of the brain gives insights about the network of physiological pathways that are active during the task. One increasingly used approach to identify causal connectivity uses the concept of Granger causality that exploits predictability of activity in one region by past activity in other regions of the brain. Owing to the complexity of the data, selecting components for the analysis of causality as a preprocessing step has to be performed. This includes predetermined—and often arbitrary—exclusion of information. Therefore, the system is confounded by latent sources. In this paper, the effect of latent confounders is demonstrated, and paths of influence among three components are studied. While methods for analysing Granger causality are commonly based on linear vector autoregressive models, the effects of latent confounders are expected to be present also in nonlinear systems. Therefore, all analyses are also performed for a simulated nonlinear system and discussed with regard to applications in neuroscience.
Original languageEnglish
Article number20110612
JournalPhilosophical Transactions of the Royal Society A: Mathematical, Physical & Engineering Sciences
Volume371
Issue number1997
Early online date15 Jul 2013
DOIs
Publication statusPublished - 28 Aug 2013

Fingerprint

neurology
network analysis
Granger Causality
Directed Network
Neuroscience
Network Analysis
Electric network analysis
nonlinear systems
brain
Nonlinear systems
Brain
Nonlinear Systems
Vector Autoregressive Model
Predictability
preprocessing
Causality
Preprocessing
Pathway
Connectivity
Path

Keywords

  • Granger causality
  • renormalized partial directed coherence
  • vector autoregressive modelling
  • latent confounders

Cite this

The impact of latent confounders in directed network analysis in neuroscience. / Ramb, Rebecca (Corresponding Author); Eichler, Michael; Ing, Alex; Thiel, Marco; Weiller, Cornelius; Grebogi, Celso; Schwarzbauer, Christian; Timmer, Jens; Schelter, Bjoern .

In: Philosophical Transactions of the Royal Society A: Mathematical, Physical & Engineering Sciences, Vol. 371, No. 1997, 20110612, 28.08.2013.

Research output: Contribution to journalArticle

@article{896123829a80442e8058e3fe7f0195e3,
title = "The impact of latent confounders in directed network analysis in neuroscience",
abstract = "In the analysis of neuroscience data, the identification of task-related causal relationships between various areas of the brain gives insights about the network of physiological pathways that are active during the task. One increasingly used approach to identify causal connectivity uses the concept of Granger causality that exploits predictability of activity in one region by past activity in other regions of the brain. Owing to the complexity of the data, selecting components for the analysis of causality as a preprocessing step has to be performed. This includes predetermined—and often arbitrary—exclusion of information. Therefore, the system is confounded by latent sources. In this paper, the effect of latent confounders is demonstrated, and paths of influence among three components are studied. While methods for analysing Granger causality are commonly based on linear vector autoregressive models, the effects of latent confounders are expected to be present also in nonlinear systems. Therefore, all analyses are also performed for a simulated nonlinear system and discussed with regard to applications in neuroscience.",
keywords = "Granger causality, renormalized partial directed coherence, vector autoregressive modelling, latent confounders",
author = "Rebecca Ramb and Michael Eichler and Alex Ing and Marco Thiel and Cornelius Weiller and Celso Grebogi and Christian Schwarzbauer and Jens Timmer and Bjoern Schelter",
year = "2013",
month = "8",
day = "28",
doi = "10.1098/rsta.2011.0612",
language = "English",
volume = "371",
journal = "Philosophical Transactions of the Royal Society A: Mathematical, Physical & Engineering Sciences",
issn = "1364-503X",
publisher = "Royal Society Publishing",
number = "1997",

}

TY - JOUR

T1 - The impact of latent confounders in directed network analysis in neuroscience

AU - Ramb, Rebecca

AU - Eichler, Michael

AU - Ing, Alex

AU - Thiel, Marco

AU - Weiller, Cornelius

AU - Grebogi, Celso

AU - Schwarzbauer, Christian

AU - Timmer, Jens

AU - Schelter, Bjoern

PY - 2013/8/28

Y1 - 2013/8/28

N2 - In the analysis of neuroscience data, the identification of task-related causal relationships between various areas of the brain gives insights about the network of physiological pathways that are active during the task. One increasingly used approach to identify causal connectivity uses the concept of Granger causality that exploits predictability of activity in one region by past activity in other regions of the brain. Owing to the complexity of the data, selecting components for the analysis of causality as a preprocessing step has to be performed. This includes predetermined—and often arbitrary—exclusion of information. Therefore, the system is confounded by latent sources. In this paper, the effect of latent confounders is demonstrated, and paths of influence among three components are studied. While methods for analysing Granger causality are commonly based on linear vector autoregressive models, the effects of latent confounders are expected to be present also in nonlinear systems. Therefore, all analyses are also performed for a simulated nonlinear system and discussed with regard to applications in neuroscience.

AB - In the analysis of neuroscience data, the identification of task-related causal relationships between various areas of the brain gives insights about the network of physiological pathways that are active during the task. One increasingly used approach to identify causal connectivity uses the concept of Granger causality that exploits predictability of activity in one region by past activity in other regions of the brain. Owing to the complexity of the data, selecting components for the analysis of causality as a preprocessing step has to be performed. This includes predetermined—and often arbitrary—exclusion of information. Therefore, the system is confounded by latent sources. In this paper, the effect of latent confounders is demonstrated, and paths of influence among three components are studied. While methods for analysing Granger causality are commonly based on linear vector autoregressive models, the effects of latent confounders are expected to be present also in nonlinear systems. Therefore, all analyses are also performed for a simulated nonlinear system and discussed with regard to applications in neuroscience.

KW - Granger causality

KW - renormalized partial directed coherence

KW - vector autoregressive modelling

KW - latent confounders

U2 - 10.1098/rsta.2011.0612

DO - 10.1098/rsta.2011.0612

M3 - Article

VL - 371

JO - Philosophical Transactions of the Royal Society A: Mathematical, Physical & Engineering Sciences

JF - Philosophical Transactions of the Royal Society A: Mathematical, Physical & Engineering Sciences

SN - 1364-503X

IS - 1997

M1 - 20110612

ER -