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 Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

A multimodal evaluation of transcranial photobiomodulation in mild cognitive impairment: Cognitive, metabolic, and neuroimaging outcomes of a pilot randomized controlled trial.

Journal of Alzheimer's disease : JAD·2026
Same author

Genetic and Hormonal Contributions to Psychosis Symptoms in Alzheimer's Disease: A Sex-Stratified Analysis.

The American journal of geriatric psychiatry : official journal of the American Association for Geriatric Psychiatry·2026
Same author

Post COVID-19 condition is associated with altered regional cerebral blood volume as revealed by dynamic susceptibility contrast MRI.

Frontiers in neuroimaging·2026
Same author

Evaluating Tests of Cognition using a Computerized Touch-Sensitive Tablet, Eye Tracking, and Functional Magnetic Resonance Imaging.

Journal of visualized experiments : JoVE·2026
Same author

Clinical Manifestations.

Alzheimer's & dementia : the journal of the Alzheimer's Association·2025
Same author

Dementia Care Research and Psychosocial Factors.

Alzheimer's & dementia : the journal of the Alzheimer's Association·2025
Same journal

Benchmarking fMRI Denoising Pipelines.

Human brain mapping·2026
Same journal

Modeled Long-Term Effects of Psilocybin on Dynamic Activity and Effective Connectivity of Fronto-Striatal-Thalamic Circuits.

Human brain mapping·2026
Same journal

Intrinsic Functional Architecture Reflects Individual Differences in Passive Working Memory: An Exploratory Resting-State fMRI Study.

Human brain mapping·2026
Same journal

Symptom Overlap and Neurobiological Similarities Between Posttraumatic Stress Disorder and Tinnitus.

Human brain mapping·2026
Same journal

Test-Retest Reliability of Sensorimotor Activity Measured With Spinal Cord fMRI.

Human brain mapping·2026
Same journal

The Human Visual Claustrum Responses to Physical Stimulus Properties and Subjective Content During Movie Viewing.

Human brain mapping·2026
See all related articles

Related Experiment Video

Updated: May 2, 2026

A Protocol for the Administration of Real-Time fMRI Neurofeedback Training
07:05

A Protocol for the Administration of Real-Time fMRI Neurofeedback Training

Published on: August 24, 2017

13.7K

Comparing within-subject classification and regularization methods in fMRI for large and small sample sizes.

Nathan W Churchill1, Grigori Yourganov, Stephen C Strother

  • 1Rotman Research Institute, Baycrest Hospital, Toronto, Ontario, Canada; Department of Medical Biophysics, University of Toronto, Toronto, Ontario, Canada.

Human Brain Mapping
|March 19, 2014
PubMed
Summary
This summary is machine-generated.

The choice of regularizer, not the classifier, significantly impacts functional magnetic resonance imaging (fMRI) analysis. Principal Component Analysis (PCA) enhances reproducibility, while Lᵖ-norms prioritize prediction accuracy in within-subject fMRI studies.

Keywords:
BOLD fMRIdata-driven metricshead motionmodel optimizationmultivariate analysisphysiological noisepreprocessing

More Related Videos

Simultaneous Data Collection of fMRI and fNIRS Measurements Using a Whole-Head Optode Array and Short-Distance Channels
08:19

Simultaneous Data Collection of fMRI and fNIRS Measurements Using a Whole-Head Optode Array and Short-Distance Channels

Published on: October 20, 2023

2.0K
Large-scale Reconstructions and Independent, Unbiased Clustering Based on Morphological Metrics to Classify Neurons in Selective Populations
12:27

Large-scale Reconstructions and Independent, Unbiased Clustering Based on Morphological Metrics to Classify Neurons in Selective Populations

Published on: February 15, 2017

6.3K

Related Experiment Videos

Last Updated: May 2, 2026

A Protocol for the Administration of Real-Time fMRI Neurofeedback Training
07:05

A Protocol for the Administration of Real-Time fMRI Neurofeedback Training

Published on: August 24, 2017

13.7K
Simultaneous Data Collection of fMRI and fNIRS Measurements Using a Whole-Head Optode Array and Short-Distance Channels
08:19

Simultaneous Data Collection of fMRI and fNIRS Measurements Using a Whole-Head Optode Array and Short-Distance Channels

Published on: October 20, 2023

2.0K
Large-scale Reconstructions and Independent, Unbiased Clustering Based on Morphological Metrics to Classify Neurons in Selective Populations
12:27

Large-scale Reconstructions and Independent, Unbiased Clustering Based on Morphological Metrics to Classify Neurons in Selective Populations

Published on: February 15, 2017

6.3K

Area of Science:

  • Neuroimaging
  • Machine Learning
  • Statistical Analysis

Background:

  • Multivariate classifiers are increasingly used for functional magnetic resonance imaging (fMRI) data analysis.
  • Regularization is crucial for stabilizing high-dimensional fMRI models, but classifier-regularizer interactions require further investigation.
  • Existing studies primarily compare classifiers on multisubject datasets, leaving performance in within-subject analyses unclear.

Purpose of the Study:

  • To evaluate the performance of different classifier and regularizer combinations for within-subject fMRI analyses.
  • To assess the impact of signal strength and sample size on classifier/regularizer model performance.
  • To compare prediction accuracy and spatial reproducibility across various methods.

Main Methods:

  • Compared four classifiers: Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), Logistic Regression, and Support Vector Machines (SVM).
  • Utilized four regularizers: Principal Component Analysis (PCA), Independent Component Analysis (ICA), L₁-norm, and L₂-norm penalization on kernel features.
  • Evaluated prediction accuracy (P) and spatial reproducibility (R) on single-subject fMRI data across different task contrasts and sample sizes (6-96 scans).

Main Results:

  • Regularizer choice had a greater impact on signal detection than the classifier model.
  • PCA maximized reproducibility and signal-to-noise ratio (SNR), while Lᵖ-norms maximized prediction accuracy.
  • ICA resulted in low reproducibility and classifier-dependent prediction accuracy.
  • Trade-offs between prediction and reproducibility varied with optimization criteria, with PCA offering the widest range.
  • Findings were consistent across task contrasts, sample sizes, and region-of-interest (ROI)-based analyses.

Conclusions:

  • Regularization strategy is a critical determinant of success in within-subject fMRI classification.
  • PCA is optimal for maximizing spatial reproducibility, whereas Lᵖ-norms are better for prediction accuracy.
  • The choice of method should be guided by the specific goals of the fMRI analysis (e.g., prediction vs. reproducibility).