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 computer-aided MFCC-based HMM system for automatic auscultation.

Sunita Chauhan1, Ping Wang, Chu Sing Lim

  • 1School of Mechanical and Aerospace Engineering, Nanyang Technological University, Singapore 639798, Singapore. mcsunita@ntu.edu.sg

Computers in Biology and Medicine
|November 30, 2007
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

Bactericidal antibiotic-phytochemical combinations against methicillin resistant Staphylococcus aureus.

Brazilian journal of microbiology : [publication of the Brazilian Society for Microbiology]·2013
Same author

Biodegradation of Low Density Polythene (LDPE) by Pseudomonas Species.

Indian journal of microbiology·2013
Same author

A Comparative Study of Metabolic Network Topology between a Pathogenic and a Non-Pathogenic Bacterium for Potential Drug Target Identification.

Summit on translational bioinformatics·2011
Same author

Maggot metabolites and their combinatory effects with antibiotic on Staphylococcus aureus.

Annals of clinical microbiology and antimicrobials·2011
Same author

Cloning and targeted disruption of two lipopolysaccharide biosynthesis genes, kdsA and waaG, of Pseudomonas aeruginosa PAO1 by site-directed mutagenesis.

Journal of molecular microbiology and biotechnology·2010
Same author

Activity and interactions of antibiotic and phytochemical combinations against Pseudomonas aeruginosa in vitro.

International journal of biological sciences·2010
Same journal

ECG arrhythmia classification via wavelet-driven feature extraction and swarm-optimised gradient boosting.

Computers in biology and medicine·2026
Same journal

Electro-osmotic metachronal cilia transport of viscoelastic blood infused with penta-hybrid nanoparticles in an oviduct: Analytical and neural network modeling.

Computers in biology and medicine·2026
Same journal

sEEGnal: an automated EEG preprocessing pipeline evaluated against expert-driven preprocessing.

Computers in biology and medicine·2026
Same journal

Corrigendum to "Integrating experimental biology, computational methods, and artificial Intelligence in anticancer drug discovery: Bridging the translational Gap" [Comput. Biol. Med. 213 (2026) 111832].

Computers in biology and medicine·2026
Same journal

Organ dose optimization for a point-of-care forearm X-ray photon-counting CT.

Computers in biology and medicine·2026
Same journal

Physics-guided transformation of breathomic feature spaces into disease-specific representations for respiratory disease classification.

Computers in biology and medicine·2026
See all related articles

This study presents a new computer-based system designed to automatically listen to and identify heart sounds. By adapting techniques used in speech recognition, the researchers created a tool that can classify normal and abnormal heartbeats with high accuracy. This technology could help healthcare providers quickly screen for heart conditions.

Area of Science:

  • Biomedical engineering research within Mel-frequency cepstral coefficient signal processing
  • Clinical diagnostics and medical informatics

Background:

Medical professionals have long relied on listening to internal body noises to assess organ health. However, manual interpretation of these acoustic signals remains subjective and prone to human error. No prior work had fully resolved the challenge of standardizing these diagnostic assessments for diverse cardiac conditions. Existing clinical practices often lack the precision required for early detection of subtle heart murmurs. This gap motivated the development of automated systems to improve diagnostic consistency. Researchers have increasingly turned to computational signal processing to address these limitations in traditional physical examinations. That uncertainty drove the exploration of advanced mathematical frameworks for acoustic pattern recognition. This investigation builds upon established signal processing methods to enhance the reliability of cardiac sound analysis.

Purpose Of The Study:

The primary aim of this study is to develop an automated system for the registration and classification of cardiac sounds. Researchers sought to address the challenges associated with the manual interpretation of internal body noises. The project focuses on creating a tool that simplifies the identification of heart signals for medical users. This effort was motivated by the need for more objective diagnostic aids in clinical environments. The authors aimed to integrate segmentation and characterization functions into a single cohesive platform. By leveraging techniques from the speech analysis domain, they intended to improve the accuracy of sound detection. The study addresses the gap in existing diagnostic tools that often struggle with the complexity of cardiac murmurs. This research ultimately strives to provide a reliable solution for screening heart conditions in primary healthcare settings.

Keywords:
automated auscultationsignal processingheart murmur detectionprobabilistic modeling

Frequently Asked Questions

The researchers propose a system that utilizes Mel-frequency cepstral coefficients to extract features, which are then processed by a hidden Markov model. This approach achieves classification rates of 95.7% for continuous murmurs, 96.25% for systolic murmurs, and 90% for diastolic murmurs.

The authors incorporate a hidden Markov model, a statistical tool originally developed for speech recognition, to categorize the acoustic patterns. This component enables the system to handle the temporal variations inherent in cardiac signals effectively.

A probabilistic comparison approach is necessary to evaluate the likelihood of different sound patterns. This technique allows the system to distinguish between normal and abnormal cardiac events by comparing input data against established statistical models.

The system processes 1381 data sets, which include both real recordings and simulated acoustic signals. This diverse data type ensures that the model is tested against a wide range of normal and abnormal heart sound profiles.

Related Experiment Videos

Main Methods:

Review approach involved designing a specialized system to capture and process heart signals. The researchers implemented segmentation functions to isolate individual cardiac events from continuous recordings. Feature extraction relied on mathematical transformations to represent the acoustic properties of the heart. The team adapted algorithms commonly used in speech analysis to interpret these complex biological signals. A hidden Markov model served as the primary architecture for classifying the extracted data. The study utilized a large collection of 1381 distinct recordings for training and validation purposes. This dataset comprised a mix of real-world clinical samples and simulated heart sounds. The final assessment employed a probabilistic comparison strategy to determine the accuracy of the automated classifications.

Main Results:

Key findings from the literature demonstrate that the system achieves high classification accuracy across multiple murmur types. The model correctly identified 96.25% of systolic murmurs within the tested dataset. Continuous murmurs showed a classification success rate of 95.7% using the proposed probabilistic approach. Diastolic murmurs were categorized with an accuracy of 90% in the final evaluation. These results indicate that the system effectively distinguishes between normal and abnormal cardiac acoustic patterns. The high performance metrics suggest that the combination of feature extraction and statistical modeling is highly reliable. The researchers observed consistent results across both real and simulated data domains. This quantitative evidence supports the utility of the system for automated cardiac sound identification.

Conclusions:

The authors suggest that their automated system demonstrates significant promise as a diagnostic support tool. Synthesis and implications indicate that high classification rates for various murmur types validate the proposed computational framework. The research highlights the effectiveness of adapting speech analysis techniques for cardiac signal interpretation. These findings imply that such technology could assist primary healthcare settings in improving patient screening efficiency. The study provides evidence that probabilistic modeling can successfully distinguish between normal and abnormal heart sounds. The authors propose that their approach offers a robust method for processing complex acoustic data. Future clinical integration could potentially reduce the burden on healthcare providers during routine physical examinations. The results confirm that the integration of specific feature extraction and modeling techniques yields reliable diagnostic outcomes.

The researchers measure the classification accuracy for three distinct murmur categories. Specifically, they report success rates of 95.7% for continuous, 96.25% for systolic, and 90% for diastolic murmurs.

The authors propose that their system holds high potential as a diagnostic aid for primary health-care sectors. They suggest this technology could streamline the identification of cardiac abnormalities during routine clinical assessments.