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 Video

Updated: Jun 11, 2026

Cross-Modal Multivariate Pattern Analysis
13:51

Cross-Modal Multivariate Pattern Analysis

Published on: November 9, 2011

Comparison of multivariate classifiers and response normalizations for pattern-information fMRI.

Masaya Misaki1, Youn Kim, Peter A Bandettini

  • 1Laboratory of Brain and Cognition, National Institute of Mental Health, National Institutes of Health, Bethesda, MD, USA. misakim@mail.nih.gov

Neuroimage
|June 29, 2010
PubMed
Summary

This study found that linear classifiers, particularly Fisher's linear discriminant and linear support vector machines, are best for decoding visual object information from fMRI data. Using t-values and selecting responsive voxels improved accuracy.

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

Group Joint ICA (gjICA): A Method for Multimodal Fusion of Concurrent EEG and fMRI Data.

Human brain mapping·2026
Same author

In-scanner thoughts contribute to resting-state functional connectivity.

Nature communications·2026
Same author

A new fMRI quality metric using multi-echo information: Theory, validation and implications.

bioRxiv : the preprint server for biology·2026
Same author

Shared Neurocardiac Pathways Linking Atrial Fibrillation and Depression: A UK Biobank Analysis.

medRxiv : the preprint server for health sciences·2026
Same author

Transformer brain encoders explain human high-level visual responses.

ArXiv·2026
Same author

Mapping high-amplitude fMRI edge time series events across space and time.

Imaging neuroscience (Cambridge, Mass.)·2026

Area of Science:

  • Neuroimaging
  • Machine Learning
  • Cognitive Neuroscience

Background:

  • Multivariate pattern analysis (MVPA) is used to decode stimulus information from fMRI data.
  • The performance of different classifiers on fMRI data is not well understood.
  • Investigating classifier performance is crucial for advancing fMRI-based decoding.

Purpose of the Study:

  • To compare the performance of six multivariate classifiers for decoding visual object categories from fMRI data.
  • To investigate the impact of response-amplitude estimates (beta vs. t-values) and pattern normalization on classification accuracy.
  • To determine optimal strategies for enhancing decoding performance in early visual and inferior temporal cortex.

Main Methods:

  • Compared six classifiers: pattern-correlation, k-nearest-neighbors, Fisher's linear discriminant, Gaussian naïve Bayes, linear SVM, and nonlinear SVM.

More Related Videos

Basics of Multivariate Analysis in Neuroimaging Data
06:35

Basics of Multivariate Analysis in Neuroimaging Data

Published on: July 24, 2010

Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data
14:27

Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data

Published on: June 26, 2013

Related Experiment Videos

Last Updated: Jun 11, 2026

Cross-Modal Multivariate Pattern Analysis
13:51

Cross-Modal Multivariate Pattern Analysis

Published on: November 9, 2011

Basics of Multivariate Analysis in Neuroimaging Data
06:35

Basics of Multivariate Analysis in Neuroimaging Data

Published on: July 24, 2010

Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data
14:27

Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data

Published on: June 26, 2013

  • Utilized BOLD fMRI data acquired with a 3T scanner using SENSE and 2-mm isotropic voxels in an event-related design.
  • Employed leave-one-stimulus-pair-out and leave-one-run-out cross-validation, and analyzed the effect of voxel selection and pattern normalization.

    • Main Results:

    • Fisher's linear discriminant and linear support vector machines demonstrated the best classification performance.
    • Using t-values instead of beta estimates for response patterns was advantageous.
    • Nonlinear classifiers performed worse than linear ones, suggesting overfitting; focused voxel selection improved decoding.

    Conclusions:

    • Linear decoders, specifically Fisher's linear discriminant and linear SVM, are recommended for decoding visual object representations using fMRI.
    • Employing t-values and selecting highly responsive voxels can significantly improve decoding accuracy.
    • The findings provide practical guidance for optimizing multivariate pattern analysis in fMRI studies.