Transfer entropy—a model-free measure of effective connectivity for the neurosciences

Max Planck Institute for Brain Research, Frankfurt, Germany.
Journal of Computational Neuroscience (Impact Factor: 2.09). 02/2011; 30(1):45-67. DOI: 10.1007/s10827-010-0262-3
Source: PubMed

ABSTRACT Understanding causal relationships, or effective connectivity, between parts of the brain is of utmost importance because a large part of the brain's activity is thought to be internally generated and, hence, quantifying stimulus response relationships alone does not fully describe brain dynamics. Past efforts to determine effective connectivity mostly relied on model based approaches such as Granger causality or dynamic causal modeling. Transfer entropy (TE) is an alternative measure of effective connectivity based on information theory. TE does not require a model of the interaction and is inherently non-linear. We investigated the applicability of TE as a metric in a test for effective connectivity to electrophysiological data based on simulations and magnetoencephalography (MEG) recordings in a simple motor task. In particular, we demonstrate that TE improved the detectability of effective connectivity for non-linear interactions, and for sensor level MEG signals where linear methods are hampered by signal-cross-talk due to volume conduction.

Download full-text


Available from: Gordon Pipa, Jul 28, 2015
1 Follower
  • Source
    • "An important property of the transfer entropy is that it does not require any particular model for the interaction between the two processes of interest (i.e., the EEG signals recorded at two electrodes). Furthermore, the transfer entropy works well when the detection of some unknown non-linear interactions is required (Vicente et al., 2011). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Alzheimer's disease (AD) is the most frequent neurodegenerative disorder and cause of dementia along aging. It is characterized by a pathological extracellular accumulation of amyloid-beta peptides that affects excitatory and inhibitory synaptic transmission. It also triggers aberrant patterns of neuronal circuit activity at the network level. Growing evidence shows that AD targets cortical neuronal networks related to cognitive functions including episodic memory and visuospatial attention. This is partially reflected by the abnormal mechanisms of cortical neural synchronization and coupling that generate resting state electroencephalographic (EEG) rhythms. The cortical neural synchronization is typically indexed by EEG power density. The EEG coupling between electrode pairs probes functional (inter-relatedness of EEG signals) and effective (casual effect from one over the other electrode) connectivity. The former is typically indexed by EEG spectral coherence (linear) or synchronization likelihood (linear-nonlinear), the latter by granger causality or information theory indexes. Here we revised resting state EEG studies in mild cognitive impairment (MCI) and AD subjects as a window on abnormalities of the cortical neural synchronization and functional and effective connectivity. Results showed abnormalities of the EEG power density at specific frequency bands (<12Hz) in the MCI and AD populations, associated to an altered functional and effective EEG connectivity among long range cortical networks (i.e. fronto-parietal and fronto-temporal). These results suggest that resting state EEG rhythms reflect the abnormal cortical neural synchronization and coupling in the brain of prodromal and overt AD subjects, possibly reflecting dysfunctional neuroplasticity of the neural transmission in long range cortical networks. Copyright © 2015. Published by Elsevier B.V.
    International journal of psychophysiology: official journal of the International Organization of Psychophysiology 02/2015; DOI:10.1016/j.ijpsycho.2015.02.008 · 2.65 Impact Factor
  • Source
    • "It does not require a model of the interaction and is inherently non-linear [4]. As a consequence, TE has been used in various applications such as identifying causal relationships between pairs of genes [7], investigating the influence of heart rate to breath rate and vice versa [6], information transfer between auditory cortical neurons using spike train data [8], exploring effective connection on MEG data associated with different types of task [4] [9], and for the localization of epileptic foci using EEG data [10] [11]. TE has not, however, been applied to more generalised cognitive processing; specifically, the degree of overall cognitive load that an individual experiences while performing a particular cognitive task. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Investigation of functional brain networks using various complex network and inferential statistical techniques in various studies have provided insight into the intricacies of the patterns of structural and functional connectivity of human brain in recent years. Most of these studies have analysed the brain networks as being undirected where the direction of information flow between various brain regions has not been considered. The directions of information flow in the functional brain networks provide additional information on how one brain region influences the other and identify influential brain regions serving as network hubs during information processing. This study used information-theoretic concept of normalized transfer entropy on the EEG data to construct directed functional brain networks during five different brain states. Using a mix of signal processing, information and graph-theoretic techniques, the findings demonstrated that directed functional brain networks constructed using normalized transfer entropy is very sensitive to the changes in the cognitive tasks and this sensitivity can be used to develop a quantitative metric to measure cognition.
    2014 ASE BigData/SocialInformatics/PASSAT/BioMedCom Conference; 12/2014
  • Source
    • "In structural health monitoring, the main objective is to locate the defects that could cause abrupt changes of the connectivity structure and adverse the performance of the system. Due to its wide range of applications, the problem of inferring causal relationships from observational data has attracted broad attention over the past few decades [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25]. "
    [Show abstract] [Hide abstract]
    ABSTRACT: Inferring the coupling structure of complex systems from time series data in general by means of statistical and information-theoretic techniques is a challenging problem in applied science. The reliability of statistical inferences requires the construction of suitable information-theoretic measures that take into account both direct and indirect influences, manifest in the form of information flows, between the components within the system. In this work, we present an application of the optimal causation entropy (oCSE) principle to identify the coupling structure of a synthetic biological system, the repressilator. Specifically, when the system reaches an equilibrium state, we use a stochastic perturbation approach to extract time series data that approximate a linear stochastic process. Then, we present and jointly apply the aggregative discovery and progressive removal algorithms based on the oCSE principle to infer the coupling structure of the system from the measured data. Finally, we show that the success rate of our coupling inferences not only improves with the amount of available data, but it also increases with a higher frequency of sampling and is especially immune to false positives.
    Entropy 11/2014; 16(6). DOI:10.3390/e16063416 · 1.56 Impact Factor
Show more