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

A Bayesian morphometry algorithm.

Edward H Herskovits1, Hanchuan Peng, Christos Davatzikos

  • 1Department of Radiology, University of Pennsylvania, 3600 Market Street, Suite 370, Room 117, Philadelphia, PA 19104, USA. ehh@ieee.org

IEEE Transactions on Medical Imaging
|June 12, 2004
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

The genetic architecture of multimodal human brain age.

Nature communications·2024
Same author

Distance-weighted Sinkhorn loss for Alzheimer's disease classification.

iScience·2024
Same author

Integrating imaging and genomic data for the discovery of distinct glioblastoma subtypes: a joint learning approach.

Scientific reports·2024
Same author

Plasma Biomarkers as Predictors of Progression to Dementia in Individuals with Mild Cognitive Impairment.

Journal of Alzheimer's disease : JAD·2024
Same author

Genetic and Clinical Correlates of AI-Based Brain Aging Patterns in Cognitively Unimpaired Individuals.

JAMA psychiatry·2024
Same author

Dimensional Neuroimaging Endophenotypes: Neurobiological Representations of Disease Heterogeneity Through Machine Learning.

ArXiv·2024

Bayesian morphological analysis effectively identifies brain structure-function relationships in MR images, even with nonlinear associations. This advanced method surpasses standard statistical tests in pinpointing brain regions linked to functional deficits.

Area of Science:

  • Neuroimaging
  • Medical Image Analysis
  • Computational Neuroscience

Background:

  • Voxel-wise statistical tests on registered MR images are common for brain structure-function analysis.
  • These methods struggle to accurately identify brain regions with nonlinear associations to clinical variables.

Purpose of the Study:

  • To introduce Bayesian morphological analysis (BMA) for MR brain image analysis.
  • To develop a model-selection framework for BMA to capture brain structure-function relationships.
  • To evaluate BMA's efficacy in identifying associations between regional brain atrophy and functional deficits.

Main Methods:

  • Bayesian networks (BNs) are used to represent probabilistic associations between brain voxels and clinical variables.
  • A model-selection framework generates a BN that models structure-function relationships from MR images and function data.

Related Experiment Videos

  • The methods are tested on two datasets: one with near-linear associations and one with nonlinear associations.

    • Main Results:

    • BMA successfully identified voxel-wise morphological changes associated with functional deficits in both linear and nonlinear datasets.
    • Standard statistical methods (t-test, paired t-test) failed to detect associations in the nonlinear dataset.
    • Bayesian networks effectively modeled complex structure-function relationships in brain imaging.

    Conclusions:

    • Bayesian morphological analysis offers a robust approach for brain structure-function analysis, particularly for nonlinear associations.
    • BMA provides a more accurate localization of structure-function relationships compared to traditional voxel-wise statistical tests.
    • This method enhances the understanding of brain morphology and its link to functional deficits in medical imaging.