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

Exact multivariate tests for brain imaging data.

Rita Almeida1, Anders Ledberg

  • 1Division of Human Brain Research, Department of Neuroscience A3:3, Karolinska Institute, Retzius väg 8, 17177 Stockholm, Sweden. Rita.Almeida@neuro.ki.se

Human Brain Mapping
|March 1, 2002
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

Otolith shape and chemistry reveal life history plasticity of an euryhaline species.

Marine environmental research·2026
Same author

Lifetime MDMA use and associations with meaning in life in the context of childhood trauma.

Scientific reports·2026
Same author

The role of social determinants in COVID-19 hospitalization disparities by migration status in Stockholm, Sweden. A population-based cohort study.

Communications medicine·2026
Same author

Atlanto-Occipital Dissociation With Catastrophic Neurological Injury: Ethical Challenges in the Intensive Care Setting.

Cureus·2026
Same author

Extracellular vesicles-associated AAVs for the treatment of Machado-Joseph disease.

Molecular therapy : the journal of the American Society of Gene Therapy·2025
Same author

Predictors of father-infant attachment security in a Portuguese sample.

Early human development·2025
Same journal

Brain Aging in Specific Phobia: An ENIGMA-Anxiety Mega-Analysis.

Human brain mapping·2026
Same journal

Talking to the Brain: Using Large Language Models as Proxies to Model Brain Semantic Features.

Human brain mapping·2026
Same journal

Emotional Context Modulates the Response to Somatosensory Stimuli Within 20 milliseconds.

Human brain mapping·2026
Same journal

GABAergic Modulation of Brain Function During Prosaccade and Antisaccade Eye Movements: Evidence From Ultra-High-Field fMRI.

Human brain mapping·2026
Same journal

Injury Severity Influences Long-Term Cognitive Control in Pediatric "Mild" Traumatic Brain Injury.

Human brain mapping·2026
Same journal

Early Adulthood Signatures of Motherhood in Brain Aging.

Human brain mapping·2026
See all related articles

This study introduces a novel statistical method for analyzing brain imaging data, like positron emission tomography (PET) and functional magnetic resonance imaging (fMRI), ensuring accurate hypothesis testing even with high-dimensional datasets. The new approach validates multivariate tests for brain imaging analysis, overcoming limitations of existing methods.

Area of Science:

  • Neuroimaging
  • Statistical analysis
  • Multivariate statistics

Background:

  • Positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) data present challenges in multivariate linear model analysis due to more variables than observations.
  • Existing methods for dimension reduction in neuroimaging data may result in non-normally distributed datasets, invalidating standard multivariate tests.
  • The theoretical foundation for null distributions in multivariate linear analysis of brain imaging data remains inadequately discussed.

Purpose of the Study:

  • To address the limitations in multivariate linear analysis for high-dimensional neuroimaging data.
  • To introduce a general method for constructing test statistics that maintain valid null distributions, irrespective of data distribution after dimension reduction.
  • To provide a theoretical basis for applying multivariate tests to brain imaging data.

Related Experiment Videos

Main Methods:

  • Developed a method for constructing test statistics based on the invariance property over a wide range of data distributions.
  • Applied the method to create a test statistic for multivariate hypotheses in PET data.
  • Utilized canonical variate analysis (CVA) to characterize significant findings.

Main Results:

  • The constructed test statistic successfully rejected the null hypothesis of no significant differences in brain activity between two conditions in a PET dataset.
  • Canonical Variate Analysis (CVA) effectively characterized the effects driving the hypothesis rejection.
  • Results from CVA showed similarity to those obtained using univariate regression analysis per voxel.

Conclusions:

  • The proposed method provides a robust framework for multivariate statistical inference in neuroimaging.
  • The invariance-based approach ensures the validity of statistical tests even after data dimension reduction.
  • The findings support the applicability and reliability of the new method for analyzing PET and fMRI data.