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

Artificial neural networks for source localization in the human brain.

U R Abeyratne1, Y Kinouchi, H Oki

  • 1Department of Electrical and Electronic Engineering, University of Tokushima, Japan.

Brain Topography
|January 1, 1991
PubMed
Summary

Neural networks offer a faster, automated solution for brain source localization, overcoming limitations of traditional iterative methods. This approach enables the use of complex head models without increased computational cost.

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

The human ether-a-go-go-related gene (hERG) current inhibition selectively prolongs action potential of midmyocardial cells to augment transmural dispersion.

Journal of physiology and pharmacology : an official journal of the Polish Physiological Society·2015
Same author

Role of innate immune sensors, TLRs, and NALP3 in rheumatoid arthritis and osteoarthritis.

Journal of long-term effects of medical implants·2015
Same author

Recurrent peritoneal inclusion cysts successfully treated with oral contraceptives: a report of two cases.

Clinical and experimental obstetrics & gynecology·2014
Same author

Reaction kinetics of nitrogen dioxide with methanol in the gas phase.

Environmental science & technology·2012
Same author

The use of arthrography in the diagnosis of temporomandibular joint synovial chondromatosis.

Dento maxillo facial radiology·2012
Same author

Efficacy of serum angiopoietin-1 measurement in the diagnosis of early rheumatoid arthritis.

Clinical and experimental rheumatology·2011

Area of Science:

  • Computational Neuroscience
  • Biomedical Engineering
  • Artificial Intelligence in Medicine

Background:

  • Brain source localization is an ill-posed problem requiring constraints on source type and head models.
  • Conventional methods rely on iterative minimization algorithms, which are computationally intensive and time-consuming.
  • Head model complexity impacts accuracy but increases computational demands and memory requirements.

Purpose of the Study:

  • To propose and evaluate the use of neural networks as an alternative to traditional minimization algorithms for brain source localization.
  • To investigate the accuracy and efficiency of neural network-based source localization.
  • To explore the potential for automating the source localization process.

Main Methods:

  • Trained a neural network using the error-backpropagation technique to compute source parameters from scalp-measured voltages.

Related Experiment Videos

  • Compared the neural network approach with conventional minimization algorithms (Simplex, Marquardt).
  • Investigated localization accuracy using extensive studies and various head models.

    • Main Results:

    • Neural networks demonstrate high feasibility as accurate source localizers.
    • Localization speed is independent of head model complexity, allowing the use of sophisticated models.
    • The method eliminates the need for initial parameter guessing, enabling automated localization.

    Conclusions:

    • Neural networks provide a rapid and efficient solution for brain source localization.
    • This approach overcomes the computational limitations associated with complex head models in traditional methods.
    • Neural networks offer a pathway to automate the entire source localization process, potentially even eliminating the need for head models if trained on experimental data.