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

Probabilistic independent component analysis for functional magnetic resonance imaging.

Christian F Beckmann1, Stephen M Smith

  • 1Medical Vision Laboratory, Department of Engineering Science and the Oxford Centre for Functional Magnetic Resonance Imaging of the Brain, University of Oxford, Oxford OX3 9DU, UK. beckmann@fmrib.ox.ac.uk

IEEE Transactions on Medical Imaging
|February 18, 2004
PubMed
Summary

This study introduces an improved probabilistic independent component analysis (ICA) for functional MRI (fMRI) data, enhancing source separation and interpretation by estimating noise and dimensionality. The method offers more accurate decomposition of brain activity and noise.

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

Consensus recommendations for clinical functional MRI applied to language mapping.

Aperture neuro·2026
Same author

Investigating the effect of increased dopamine signaling on cerebral blood flow in Major Depressive Disorder: a double-blind randomized placebo-controlled study.

Neuropsychopharmacology : official publication of the American College of Neuropsychopharmacology·2026
Same author

Multiscale heterogeneity of atypical functional connectivity in autism.

Nature. Mental health·2026
Same author

Investigating the amyloid-tau-neurodegeneration framework in Alzheimer's disease using semi-supervised multimodal imaging data fusion.

Alzheimer's & dementia (Amsterdam, Netherlands)·2026
Same author

Characterizing functional connectivity gradients for the hippocampus-amygdala complex in healthy and psychiatric cohorts.

Brain structure & function·2026
Same author

Social dysfunction and default mode network functional integrity in neuropsychiatric disorders: A cross-disorder replication and generalization study.

European neuropsychopharmacology : the journal of the European College of Neuropsychopharmacology·2026

Area of Science:

  • Neuroimaging
  • Computational Neuroscience
  • Signal Processing

Background:

  • Functional MRI (fMRI) data analysis often relies on Independent Component Analysis (ICA).
  • Standard ICA methods face challenges with nonsquare mixing and accurate noise estimation in fMRI.
  • Overfitting and interpretation issues can arise from unaddressed noise and dimensionality in fMRI data.

Purpose of the Study:

  • To present an integrated probabilistic ICA approach for fMRI data.
  • To enable nonsquare mixing analysis and robust noise estimation.
  • To improve the interpretability and accuracy of independent components in fMRI.

Main Methods:

  • Developed a probabilistic ICA framework for fMRI data.
  • Employed Bayesian analysis for objective Gaussian noise estimation and true dimensionality determination.

Related Experiment Videos

  • Incorporated temporal prewhitening and variance normalization for enhanced signal processing.
  • Utilized prior information on spatiotemporal source characteristics and Gaussian mixture models for inference.

    • Main Results:

    • Achieved an asymptotically unique decomposition of fMRI data.
    • Reduced interpretation ambiguity by associating components with single physical/physiological processes.
    • Demonstrated improved performance on real and artificial fMRI data compared to classical ICA and GLM.
    • Highlighted the utility of dimensionality reduction techniques for weak activation detection.

    Conclusions:

    • The proposed probabilistic ICA offers a more robust and interpretable method for fMRI data analysis.
    • Objective noise estimation and dimensionality assessment are crucial for accurate source separation.
    • The integrated approach enhances the spatio-temporal accuracy of identified neural components in fMRI studies.