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

Bayesian inverse analysis of neuromagnetic data using cortically constrained multiple dipoles.

Toni Auranen1, Aapo Nummenmaa, Matti S Hämäläinen

  • 1Laboratory of Computational Engineering, Helsinki University of Technology, Espoo, Finland. Toni.Auranen@hut.fi

Human Brain Mapping
|March 21, 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

Standardizing TMS Intensity Across Different Coils Using Individualized Electric Field Modeling.

Human brain mapping·2026
Same author

Connectivity-guided transcranial magnetic stimulation of the dopamine system: A proof-of-concept study.

Brain stimulation·2026
Same author

Conductivity Deviations as Virtual Sources in Magnetoencephalography.

Brain topography·2026
Same author

Pathfinder: Parallel quasi-Newton variational inference.

Journal of machine learning research : JMLR·2026
Same author

Electric and magnetic field characteristics of the Re5 brain stimulation system.

Brain stimulation·2026
Same author

A Digital Anatomical Atlas of the Human Cerebellum at Subfolial Resolution.

Human brain mapping·2026

This study enhances a Bayesian model for magnetoencephalography (MEG) to improve dipole localization using individual brain anatomy. The improved model offers better data fitting but requires adjustments for efficient dipole mode switching.

Area of Science:

  • Neuroimaging
  • Computational Neuroscience
  • Biophysics

Background:

  • Magnetoencephalography (MEG) is a neuroimaging technique used to measure magnetic fields produced by electrical activity in the brain.
  • Accurate source localization of neural activity is crucial for understanding brain function.
  • Previous Bayesian models have shown promise in localizing neural sources but require further refinement for clinical application.

Purpose of the Study:

  • To enhance an existing Bayesian model for MEG data analysis by incorporating individual subject brain surface reconstructions.
  • To improve the efficiency and accuracy of dipole localization by introducing cortical location and orientation constraints.
  • To address limitations in dipole mode switching within the model for better posterior distribution analysis.

Main Methods:

Related Experiment Videos

  • Adapted a Bayesian model for magnetoencephalographic (MEG) data analysis.
  • Incorporated individual brain surface reconstructions with cortical constraints.
  • Utilized spherical angle coordinates for efficient Markov chain Monte Carlo sampling.
  • Simplified the likelihood function using singular value decomposition (SVD) of the gain matrix.
  • Analyzed simulated and empirical MEG data from auditory and visual stimulation.

Main Results:

  • The enhanced model produced reasonable source localization solutions and adequate data fits with minimal manual intervention.
  • Rigid cortical constraints presented challenges for efficient dipole mode switching, particularly with multimodal posterior distributions.
  • The model's efficiency in sampling dipole locations was improved through SVD-based data simplification.

Conclusions:

  • The developed Bayesian model offers a viable approach for MEG source localization, integrating individual anatomy.
  • Further refinement is needed to overcome limitations related to rigid cortical constraints and improve dipole mode switching.
  • Future directions include exploring loose orientation constraints and hybrid prelocalization models for enhanced performance.