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 wavelet-based optimal texture feature set for classification of brain tumours.

M Sasikala1, N Kumaravel

  • 1Department of Instrumentation Engineering, Madras Institute of Technology, Anna University, Chennai, Tamil Nadu, India. sasi_yugesh@yahoo.com

Journal of Medical Engineering & Technology
|April 25, 2008
PubMed
Summary

This study introduces optimal texture features for classifying brain tumors in magnetic resonance images (MRIs). A genetic algorithm achieved 98% accuracy using only four features, outperforming other methods.

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 flexible real-time surface electromyography-based wheelchair control system using upper arm or neck muscles.

Disability and rehabilitation. Assistive technology·2026
Same author

Molecular characterization of virulent genotype XIII Newcastle disease virus isolates from desi-chickens in Namakkal, South India.

Veterinary research communications·2026
Same author

Smartphone-Based Anemia Screening <i>via</i> Conjunctival Imaging with 3D-Printed Spacer: A Cost-Effective Geospatial Health Solution.

Current medical imaging·2025
Same author

Development of a High-Throughput Microscope for the Analysis of Peripheral Blood Smears for Anemia Screening.

Journal of biophotonics·2025
Same author

The role of beat-by-beat cardiac features in machine learning classification of ischemic heart disease (IHD) in magnetocardiogram (MCG).

Biomedical physics & engineering express·2024
Same author

A method for noninvasive beat-by-beat visualization of His bundle signals.

Annals of noninvasive electrocardiology : the official journal of the International Society for Holter and Noninvasive Electrocardiology, Inc·2023

Area of Science:

  • Medical Imaging
  • Artificial Intelligence
  • Biomedical Engineering

Background:

  • Accurate brain tumor classification from magnetic resonance images (MRIs) is crucial for diagnosis and treatment.
  • Traditional methods often require extensive feature sets, impacting computational efficiency.

Purpose of the Study:

  • To investigate the efficacy of optimal texture features for classifying normal brain, benign tumors, and malignant tumors using MRIs.
  • To compare the performance of a genetic algorithm-based feature selection method against principal component analysis and classical sequential methods.

Main Methods:

  • A wavelet-based texture feature set was derived from regions of interest in brain MRIs.
  • An artificial neural network classifier was employed to evaluate feature performance.

Related Experiment Videos

  • A genetic algorithm was used for optimal feature selection, comparing its efficiency against principal component analysis and sequential methods.

    • Main Results:

    • The genetic algorithm achieved a high classification performance of 98% using only four texture features out of 29.
    • Principal component analysis and classical sequential methods required a larger feature set to reach the same 98% accuracy.
    • Optimal texture features identified by the genetic algorithm, including specific measures of angular second moment, sum variance, and information measure of correlation II, demonstrated superior classification performance.

    Conclusions:

    • Optimal texture features, selected by a genetic algorithm, provide a highly accurate and computationally efficient method for brain tumor classification in MRIs.
    • The genetic algorithm approach offers a significant advantage in reducing the number of features required for high diagnostic accuracy compared to traditional methods.
    • This approach holds promise for improving automated brain tumor detection and classification systems.