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

Characterizing the response of PET and fMRI data using multivariate linear models

K J Worsley1, J B Poline, K J Friston

  • 1Department of Mathematics and Statistics, McGill University, 805 Sherbrooke Street West, Montreal, Québec, H3A 2K6, Canada.

Neuroimage
|February 7, 1998
PubMed
Summary

This study introduces a new multivariate linear model (MLM) method to analyze brain responses in PET and fMRI scans. The method effectively captures correlations between brain activity and external predictors, improving upon existing Canonical Variates Analysis (CVA) techniques.

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

Bridging the neuro-AI chasm: a framework for scalable, contextually adaptive training resources in large-scale brain data science.

Frontiers in psychology·2026
Same author

The Neuroimaging Data Model Linear Regression Tool (nidm_linreg): PyNIDM Project.

F1000Research·2024
Same author

Automatic rating of incomplete hippocampal inversions evaluated across multiple cohorts.

ArXiv·2024
Same author

Crystallization of ApoA1 and ApoE4 nanolipoprotein particles and initial XFEL-based structural studies.

Crystals·2022
Same author

Correction to: A Standards Organization for Open and FAIR Neuroscience: the International Neuroinformatics Coordinating Facility.

Neuroinformatics·2021
Same author

A Standards Organization for Open and FAIR Neuroscience: the International Neuroinformatics Coordinating Facility.

Neuroinformatics·2021

Area of Science:

  • Neuroimaging
  • Biostatistics
  • Cognitive Neuroscience

Background:

  • Characterizing brain responses in neuroimaging is crucial for understanding cognitive processes.
  • Existing methods for analyzing PET and fMRI data have limitations in capturing complex temporal correlations.

Purpose of the Study:

  • To present a novel multivariate linear model (MLM) method for brain response characterization in PET and fMRI.
  • To capture correlations between neuroimaging scans and external predictor variables.

Main Methods:

  • Utilizes Canonical Variates Analysis (CVA) on estimated effects from a multivariate linear model (MLM).
  • Incorporates temporal correlations, making the method suitable for both fMRI and PET data.
  • Employs standard multivariate statistics for correlation testing and inference on canonical variates.

Related Experiment Videos

Main Results:

  • The MLM-CVA method successfully incorporates temporal correlations, enhancing suitability for fMRI and PET.
  • Standard statistical tests can be used for inference, reducing reliance on simulations.
  • Application to an fMRI dataset revealed response patterns more aligned with prior non-CVA analyses.

Conclusions:

  • The proposed MLM-CVA method offers an advancement in analyzing brain responses from PET and fMRI data.
  • This approach provides a statistically robust framework for identifying relationships between neural activity and experimental conditions.
  • The method demonstrates improved sensitivity in detecting response patterns compared to other CVA techniques.