University of Sussex
Browse
seth18488_preprint.pdf (341.81 kB)

Causal connectivity of evolved neural networks during behavior

Download (341.81 kB)
journal contribution
posted on 2023-06-08, 08:18 authored by Anil SethAnil Seth
To show how causal interactions in neural dynamics are modulated by behavior, it is valuable to analyze these interactions without perturbing or lesioning the neural mechanism. This paper proposes a method, based on a graph-theoretic extension of vector autoregressive modeling and 'Granger causality,' for characterizing causal interactions generated within intact neural mechanisms. This method, called 'causal connectivity analysis' is illustrated via model neural networks optimized for controlling target fixation in a simulated head-eye system, in which the structure of the environment can be experimentally varied. Causal connectivity analysis of this model yields novel insights into neural mechanisms underlying sensorimotor coordination. In contrast to networks supporting comparatively simple behavior, networks supporting rich adaptive behavior show a higher density of causal interactions, as well as a stronger causal flow from sensory inputs to motor outputs. They also show different arrangements of 'causal sources' and 'causal sinks': nodes that differentially affect, or are affected by, the remainder of the network. Finally, analysis of causal connectivity can predict the functional consequences of network lesions. These results suggest that causal connectivity analysis may have useful applications in the analysis of neural dynamics.

History

Publication status

  • Published

File Version

  • Accepted version

Journal

Network: Computation in Neural Systems

ISSN

0954-898X

Publisher

Informa Healthcare

Issue

1

Volume

16

Page range

35-54

Pages

20.0

Department affiliated with

  • Informatics Publications

Notes

Originality: Introduces a novel method for characterizing causal interactions in neural networks without introducing lesions or artificial perturbations. Rigour: Combines graph theory with econometric time series analysis to provide a novel interdisciplinary methodology for deriving 'causal graphs' of neural dynamics. Illustrated the method by application to an embodied neuronal simulation, showing that causal pathways can be modulated by adaptation to rich environments. Significance: The proposed method provides a very general technique for analyzing network dynamics, and the illustrative example shows its usefulness for analyzing embodied and environmentally embedded neural systems, a key challenge for current neuroscience. Impact: The associated MatLab toolbox has been downloaded more than fifty times and is in increasingly wide use both in the neuroscience community and in associated communities such as financial economics and bioinformatics. As the importance of analyzing causal interactions is more widely appreciated, the paper and the methods described therein should become more widely disseminated and extended.

Full text available

  • Yes

Peer reviewed?

  • Yes

Legacy Posted Date

2012-02-06

First Open Access (FOA) Date

2016-03-22

First Compliant Deposit (FCD) Date

2016-11-16

Usage metrics

    University of Sussex (Publications)

    Categories

    No categories selected

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC