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

Fuzzy K-nearest neighbor classifiers for ventricular arrhythmia detection.

D Cabello1, S Barro, J M Salceda

  • 1Departamento de Electronica, Facultad de Fisica, Universidad de Santiago de Compostela, Spain.

International Journal of Bio-Medical Computing
|February 1, 1991
PubMed
Summary

This study evaluated four classifiers for detecting ventricular arrhythmias in ECGs. Fuzzy Covariance clustering with K-nearest-neighbor achieved the best performance, showing high efficiency in arrhythmia detection.

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 method for designing birdcage coils based on a simplified magnetic field model, validated experimentally at 4 T, 7 T, and 15.2 T.

HardwareX·2026
Same author

An unusual type of early gastric cancer with aggressive behavior: It is not always adenocarcinoma!

Revista de gastroenterologia de Mexico (English)·2025
Same author

Drug screening targeting TREM2-TYROBP transmembrane binding.

Molecular medicine (Cambridge, Mass.)·2025
Same author

Modification of Quark-Gluon Distributions in Nuclei by Correlated Nucleon Pairs.

Physical review letters·2024
Same author

Promoting second victims' intervention support programs.

Journal of healthcare quality research·2024
Same author

Pollen production in olive cultivars and its interannual variability.

Annals of botany·2023

Area of Science:

  • Biomedical Engineering
  • Signal Processing
  • Machine Learning

Background:

  • Ventricular arrhythmias require accurate detection for timely intervention.
  • Electrocardiogram (ECG) analysis is crucial for diagnosing cardiac conditions.
  • Automated classification of ECGs can aid clinical decision-making.

Purpose of the Study:

  • To compare the efficiency of four distinct classification algorithms for detecting ventricular arrhythmias in ECG traces.
  • To evaluate the performance of K-nearest-neighbor and single-nearest-prototype algorithms when parameterized by Fuzzy C-Means and Fuzzy Covariance clustering.
  • To determine the optimal parameters for these classifiers to minimize classification errors.

Main Methods:

  • Utilized Principal Component Analysis (PCA) and cardiologist classification to define 5 ECG trace classes.

Related Experiment Videos

  • Employed K-nearest-neighbor and single-nearest-prototype algorithms, parameterized by Fuzzy C-Means and Fuzzy Covariance clustering.
  • Developed multiple learning and test sets from 90 ECG traces to train and validate classifiers.
  • Optimized classifier parameters (K for KNN, cluster volumes for Fuzzy Covariance) through iterative testing.

    • Main Results:

    • Fuzzy Covariance clustering demonstrated superior perception of cluster structure compared to Fuzzy C-Means.
    • The K-nearest-neighbor algorithm, when parameterized by Fuzzy Covariance, achieved an overall empirical error ratio of 0.10.
    • This combination significantly outperformed other tested classifier-clustering pairings in arrhythmia detection.

    Conclusions:

    • Fuzzy Covariance clustering is a more effective method for parameterizing classifiers in ECG analysis.
    • The K-nearest-neighbor algorithm, enhanced by Fuzzy Covariance, shows significant promise for accurate ventricular arrhythmia detection.
    • This approach offers a robust and efficient tool for automated ECG interpretation.