Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

A Metropolis-Hastings algorithm for dynamic causal models.

Justin R Chumbley1, Karl J Friston, Tom Fearn

  • 1Wellcome Centre for Neuroimaging, Institute of Neurology, UCL, 12 Queen Square, London, WC1N 3BG, UK. j.chumbley@fil.ion.ucl.ac.uk

Neuroimage
|September 22, 2007
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Indirect comparison of the immunogenicity of 15-valent and 20-valent pneumococcal conjugate vaccines in children using a 2+1 schedule.

Expert review of vaccines·2025
Same author

Reliability of dynamic causal modelling of resting-state magnetoencephalography.

Human brain mapping·2024
Same author

Narrative as active inference: an integrative account of cognitive and social functions in adaptation.

Frontiers in psychology·2024
Same author

Forgetting ourselves in flow: an active inference account of flow states and how we experience ourselves within them.

Frontiers in psychology·2024
Same author

Two distinct networks for encoding goals and forms of action: An effective connectivity study.

Proceedings of the National Academy of Sciences of the United States of America·2024
Same author

Inferring Trajectories of Psychotic Disorders Using Dynamic Causal Modeling.

Computational psychiatry (Cambridge, Mass.)·2024

Dynamic causal modelling (DCM) is a method for brain network analysis. This study shows that common approximations in DCM are valid for practical use and that results are not sensitive to prior assumptions.

Area of Science:

  • Neuroscience
  • Computational Neuroscience
  • Systems Neuroscience

Background:

  • Dynamic Causal Modelling (DCM) infers brain network causal architecture.
  • Current DCM formulations rely on normality assumptions for Bayesian inversion.
  • These assumptions simplify posterior distribution estimation but may lack analytical rigor.

Purpose of the Study:

  • Critique and numerically evaluate the normality assumptions in DCM.
  • Assess the performance of conventional DCM inversion schemes.
  • Investigate the sensitivity of DCM posteriors to prior specifications.

Main Methods:

  • Employed the Metropolis-Hastings algorithm for Bayesian inversion, avoiding normality assumptions.
  • Compared approximate (Laplace) posteriors with exact (MCMC) posteriors using conventional priors.

Related Experiment Videos

  • Evaluated exact posteriors with both conventional and uninformative priors.

    • Main Results:

    • The Laplace approximation for DCM posterior distributions is suitable for practical applications.
    • DCM posterior inferences are largely insensitive to the choice of prior distributions.
    • Numerical evaluations confirmed the validity of commonly used DCM assumptions.

    Conclusions:

    • Conventional normality assumptions in DCM are practically valid.
    • The choice of priors has minimal impact on DCM results.
    • The Metropolis-Hastings approach provides a robust method for DCM analysis.