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

Improving tissue classification in MRI: a three-dimensional multispectral discriminant analysis method with automated

G Harris1, N C Andreasen, T Cizadlo

  • 1Mental Health Clinical Research Center, University of Iowa College of Medicine and Hospitals and Clinics, Iowa City 52242, USA.

Journal of Computer Assisted Tomography
|March 2, 1999
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

Prevalence of structural and idiopathic epilepsy in brachycephalic and non-brachycephalic dogs in the context of the International Veterinary Epilepsy Task Force guidelines.

The Journal of small animal practice·2025
Same author

UK National Fingerprint Collaborative Exercise 2022-23.

Science & justice : journal of the Forensic Science Society·2024
Same author

Automated detection, delineation and quantification of whole-body bone metastasis using FDG-PET/CT images.

Physical and engineering sciences in medicine·2023
Same author

Letter: budesonide for functional dyspepsia with duodenal eosinophilia-randomised, double-blind, placebo-controlled parallel-group trial.

Alimentary pharmacology & therapeutics·2021
Same author

Rapid Access Palliative Radiotherapy Programmes.

Clinical oncology (Royal College of Radiologists (Great Britain))·2020
Same author

Schistosoma mansoni cathepsin D1: Biochemical and biophysical characterization of the recombinant enzyme expressed in HEK293T cells.

Protein expression and purification·2019

An automated method for brain tissue classification using multispectral MRI sequences significantly improved reliability and accuracy over manual methods. This approach offers a more efficient and valid way to segment gray matter, white matter, and cerebrospinal fluid.

Area of Science:

  • Neuroimaging
  • Medical Image Analysis
  • Biomedical Engineering

Background:

  • Accurate tissue classification in multispectral MRI (T1, T2, PD) is crucial for quantitative analysis.
  • Operator-dependent methods for training class identification are often unreliable and computationally inefficient.
  • Partial voluming can affect the accuracy of discrete tissue classification.

Purpose of the Study:

  • To develop and validate an automated method for identifying training classes for discriminant analysis in multispectral MRI.
  • To compare the reliability and accuracy of the automated method against a supervised, operator-dependent approach.
  • To implement a fuzzy classification method to correct for partial voluming effects.

Main Methods:

  • Acquired co-registered multispectral MRI (T1, T2, PD) on a 1.5 T scanner.

Related Experiment Videos

  • Developed an automated method using randomly selected brain tissue samples ('plugs') to identify training classes (GM, WM, CSF) by minimizing intra-partition variance.
  • Compared automated versus operator-dependent training class identification and sharp (discrete) versus fuzzy (continuous) classification.

    • Main Results:

    • Automated sharp classification demonstrated superior interrater and intrarater reliability compared to other methods.
    • The automated method showed high validity, evidenced by excellent reproducibility (intraclass r > 0.97), sensitivity to aging, and agreement with expert manual segmentation (up to 94%).
    • Sharp automated classification slightly outperformed fuzzy classification in terms of reliability and validity metrics.

    Conclusions:

    • Automated training class identification for discriminant analysis is superior to operator-dependent methods for multispectral MRI tissue classification.
    • Sharp (discrete) classification provides slightly better results than fuzzy classification for this automated approach.
    • The developed automated multispectral discriminant analysis method is computationally efficient, reliable, and valid for segmenting brain tissues (GM, WM, CSF).