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

Contextual clustering for analysis of functional MRI data.

E Salli1, H J Aronen, S Savolainen

  • 1Laboratory of Biomedical Engineering, Helsinki University of Technology, Espoo, Finland. eero.salli@hut.fi

IEEE Transactions on Medical Imaging
|June 14, 2001
PubMed
Summary

This study introduces a new contextual clustering algorithm for analyzing functional magnetic resonance imaging (fMRI) data. The method improves neural activation detection by utilizing neighborhood information, offering better sensitivity and specificity than traditional 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

Quantifying the calcification of abdominal aorta and major side branches with deep learning.

Clinical radiology·2024
Same author

Automatic detection of Crohn's disease using quantified motility in magnetic resonance enterography: initial experiences.

Clinical radiology·2021
Same author

Enabling and promoting walking rehabilitation by paired associative stimulation after incomplete paraplegia: a case report.

Spinal cord series and cases·2020
Same author

Acceptability of flavoured pharmaceutically non-active mini-tablets in pet cats tested with a rapid 3-portal acceptance test with and without food.

Veterinary and animal science·2020
Same author

Restoration of hand function with long-term paired associative stimulation after chronic incomplete tetraplegia: a case study.

Spinal cord series and cases·2019
Same author

Difficulties in administration of oral medication formulations to pet cats: an e-survey of cat owners.

The Veterinary record·2016

Area of Science:

  • Neuroimaging
  • Statistical analysis
  • Image processing

Background:

  • Statistical parametric maps (SPM) are crucial for analyzing time-varying 3D image data, particularly in functional magnetic resonance imaging (fMRI).
  • Detecting neural activations in fMRI requires distinguishing active areas from background noise, where background intensity distribution is known but activation distributions are not.

Purpose of the Study:

  • To develop a contextual clustering algorithm for improved detection of neural activations in fMRI data.
  • To control the probability of false activations by incorporating hypothesis testing within the clustering procedure.
  • To leverage neighborhood information in statistical parametric maps, unlike traditional voxel-by-voxel methods.

Main Methods:

  • A contextual clustering procedure is applied to statistical parametric maps derived from fMRI data.

Related Experiment Videos

  • A Markov random field prior and iterated conditional modes (ICM) algorithm are utilized, with classification based on background distribution.
  • The algorithm divides SPMs into background and activation areas, performing hypothesis testing to control false positives.

    • Main Results:

    • Simulations and human fMRI experiments demonstrate superior sensitivity at a given specificity compared to voxel-by-voxel thresholding.
    • The algorithm effectively detects and delineates neural activations from noisy backgrounds in fMRI data.
    • The contextual clustering approach proves computationally efficient.

    Conclusions:

    • The developed contextual clustering algorithm enhances the detection of neural activations in fMRI.
    • Utilizing neighborhood information and background distribution improves detection accuracy over traditional methods.
    • The algorithm's efficiency and effectiveness suggest broader applications in object detection from noisy backgrounds.