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

Feature extraction for systolic heart murmur classification.

Christer Ahlstrom1, Peter Hult, Peter Rask

  • 1Department of Biomedical Engineering, University Hospital, Linköping University, IMT, SE-581 85, Linköping, Sweden. christer@imt.liu.se

Annals of Biomedical Engineering
|October 5, 2006
PubMed
Summary

An intelligent stethoscope can help classify heart murmurs. A new set of 14 features, including novel ones, achieved 86% accuracy in distinguishing pathological from physiological heart murmurs.

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

CARDIAC FINDINGS IN SWIMMING-INDUCED PULMONARY EDEMA - IMPLICATIONS FOR ACUTE ASSESSMENT.

Chest·2026
Same author

Distribution of diffuse myocardial fibrosis is uneven and associated with left ventricular function in severe aortic stenosis.

Open heart·2026
Same author

Psychiatric Disorders and Mortality in Heart Failure: A Swedish Nationwide Cohort Study.

JACC. Heart failure·2026
Same author

Implementation of sodium-glucose co-transporter 2 inhibitors use in patients with heart failure across the ejection fraction: data from the Swedish heart failure registry.

European heart journal. Quality of care & clinical outcomes·2026
Same author

An integrated Biobank in the Swedish Heart Failure Registry-clinomics, proteomics, transcriptomics and genomics.

ESC heart failure·2026
Same author

Optimizing NT-proBNP Inclusion Cut-offs for Randomized Clinical Trials in Heart Failure: Data from the Swedish Heart Failure Registry.

ESC heart failure·2026

Area of Science:

  • Cardiology
  • Biomedical Engineering
  • Signal Processing

Background:

  • Heart murmurs are critical indicators of cardiac valve disease, often detected during primary care auscultation.
  • Differentiating pathological from physiological murmurs is challenging, highlighting the need for decision support tools like an intelligent stethoscope.

Purpose of the Study:

  • To identify an optimal feature subset for the automatic classification of heart murmurs using phonocardiographic signals.
  • To evaluate the efficacy of novel and established signal processing features for murmur classification.

Main Methods:

  • Acquired phonocardiographic signals from 36 patients with aortic valve stenosis, mitral insufficiency, or physiological murmurs.
  • Extracted 207 features using techniques including Shannon energy, wavelets, fractal dimensions, and recurrence quantification analysis.

Related Experiment Videos

  • Employed Pudil's sequential floating forward selection (SFFS) to derive a 14-feature multi-domain subset for neural network classification.

    • Main Results:

    • A multi-domain feature subset of 14 features (combining existing and novel features) yielded the highest classification accuracy.
    • The selected multi-domain subset achieved 86% correct classifications, significantly outperforming single-domain sets (68% for the best runner-up).
    • The derived feature set demonstrated robustness even with noisy data.

    Conclusions:

    • The developed multi-domain feature set is superior for automatic heart murmur classification compared to single-domain approaches.
    • This feature set shows promise for enhancing the diagnostic capabilities of intelligent stethoscopes in primary healthcare.
    • The findings suggest a robust method for improving the accuracy of pathological versus physiological heart murmur identification.