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

Automatic wheeze detection based on auditory modelling.

Y Qiu1, A R Whittaker, M Lucas

  • 1Department of Mechanical Engineering, University of Glasgow, Glasgow, UK.

Proceedings of the Institution of Mechanical Engineers. Part H, Journal of Engineering in Medicine
|June 7, 2005
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

Evaluation of a new effect-site controlled, patient-maintained sedation system in dental patients.

Anaesthesia·2006
Same author

Cardiovascular changes after induction of anaesthesia.

Anaesthesia·2006
Same author

Novel targeted drugs for the treatment of multiple myeloma: from bench to bedside.

Leukemia·2005
Same author

Penetration of ceftiofur into sterile vs. Mannheimia haemolytica-infected tissue chambers in beef calves after subcutaneous administration of ceftiofur crystalline free acid sterile suspension in the ear pinna.

Journal of veterinary pharmacology and therapeutics·2005
Same author

Histone deacetylase inhibitor NVP-LAQ824 has significant activity against myeloid leukemia cells in vitro and in vivo.

Leukemia·2004
Same author

A partially blinded randomised controlled trial of patient-maintained propofol sedation and operator controlled midazolam sedation in third molar extractions.

Anaesthesia·2004
Same journal

Assessment of adaptive trabecular bone remodeling around a knee prosthesis stem using the Weinans model.

Proceedings of the Institution of Mechanical Engineers. Part H, Journal of engineering in medicine·2026
Same journal

Biomechanical evaluation of a fully cortical-threaded screw for modified cortical bone trajectory fixation: Combined experimental and finite element study.

Proceedings of the Institution of Mechanical Engineers. Part H, Journal of engineering in medicine·2026
Same journal

The influence of passive prosthetic knee joints on uneven compliant surface walking among transfemoral amputees under dual-task versus single-task conditions.

Proceedings of the Institution of Mechanical Engineers. Part H, Journal of engineering in medicine·2026
Same journal

Bone remodelling around internal fracture fixations used for treating femoral neck fractures: A finite element analysis.

Proceedings of the Institution of Mechanical Engineers. Part H, Journal of engineering in medicine·2026
Same journal

Design of a kinematic-compatible passive spinal exoskeleton with parallelepiped units.

Proceedings of the Institution of Mechanical Engineers. Part H, Journal of engineering in medicine·2026
Same journal

The tipping point: Friction and thigh length influence imminent falls from a tilting ski lift chair.

Proceedings of the Institution of Mechanical Engineers. Part H, Journal of engineering in medicine·2026
See all related articles

This study introduces a new, more accurate computer-based method for identifying wheezing sounds in patients. By mimicking how humans process sound, this tool provides a reliable, objective record of breathing issues, offering a significant improvement over traditional listening methods.

Area of Science:

  • Respiratory medicine diagnostics within auditory modelling
  • Digital signal processing for clinical wheeze detection applications

Background:

No prior work has fully resolved the limitations of automated respiratory sound analysis for clinical practice. Human auscultation remains subjective and relies heavily on the individual expertise of the clinician. That uncertainty drove the need for reliable, objective tools to assist in diagnosing airway obstruction. Prior research has shown that existing automated systems often lack the precision required for widespread adoption. These earlier attempts frequently failed to distinguish between true wheezing and other lung sounds effectively. This gap motivated the development of more sophisticated signal processing techniques. Researchers have long sought to replace manual listening with consistent, quantifiable data collection. The field currently lacks a standard, highly accurate method for identifying these specific acoustic events in patient recordings.

Purpose Of The Study:

The aim of this research is to develop an improved algorithm for the automated identification of wheezing sounds. The authors seek to overcome the limitations inherent in human auscultation, which often lacks consistency and objective documentation. This project addresses the need for a reliable tool capable of quantifying the severity of respiratory conditions. The investigators focus on creating a system that is independent of the clinician's personal experience. By utilizing auditory modelling, they intend to refine how machines interpret complex lung sounds. The study explores the effectiveness of a frequency- and duration-dependent thresholding approach for better sound classification. This work is motivated by the failure of previous automated attempts to achieve sufficient reliability for widespread clinical use. The researchers aim to provide a more precise, automated historical record of patient respiratory health.

Keywords:
respiratory sound analysissignal processingclinical diagnosticsspectrogram interpretation

Frequently Asked Questions

The researchers propose a frequency- and duration-dependent thresholding mechanism. This approach identifies wheezes by analyzing power within specific frequency bands rather than relying on total signal power, which distinguishes it from older, less precise global power methods.

The algorithm utilizes a spectrogram to visualize and mark detected wheeze components. This tool allows clinicians to see the exact timing and frequency characteristics of the sounds, providing a permanent, objective record of the patient's respiratory activity.

The authors state that analyzing power within a particular frequency range is necessary to improve detection accuracy. This technical requirement prevents the system from being overwhelmed by background noise, unlike previous methods that used global power thresholds.

Related Experiment Videos

Main Methods:

The review approach involved developing a frequency- and duration-dependent threshold algorithm to process respiratory sounds. Researchers designed this system to mimic human sound perception through advanced signal processing techniques. They implemented a strategy that evaluates power within specific frequency bands rather than using global signal metrics. The team tested this software on a cohort of 36 human subjects. Eleven of these participants displayed clinical signs of wheezing during the recording sessions. The investigators automatically extracted the mean frequency and duration for every identified sound component. They utilized spectrograms to visually annotate and verify the presence of these acoustic events. This systematic evaluation allowed for a direct performance comparison against older, less effective diagnostic models.

Main Results:

Key findings from the literature indicate that the new algorithm achieves a marked improvement in accuracy compared to previous diagnostic tools. The system successfully identified wheezing characteristics in 11 out of 36 tested subjects. By shifting the threshold logic to frequency-specific power, the researchers reduced errors common in earlier models. This refined approach allows for the precise, automated extraction of both frequency and duration metrics for every detected event. The data show that this method provides a more reliable alternative to traditional human auscultation. The results confirm that the algorithm effectively marks wheezes on spectrograms for clear clinical interpretation. This performance gain highlights the effectiveness of incorporating auditory-inspired logic into respiratory sound analysis. The study provides evidence that this specific thresholding technique enhances the overall sensitivity of automated detection systems.

Conclusions:

The authors propose that their novel thresholding approach significantly enhances the reliability of automated respiratory sound monitoring. This synthesis suggests that moving away from global power metrics improves detection sensitivity. The findings imply that frequency-specific analysis provides a more robust framework for identifying wheezing components. Their results demonstrate that this method outperforms previous diagnostic algorithms in clinical testing scenarios. The researchers believe this tool offers a clear path toward objective, experience-independent assessment of airway conditions. By marking events on spectrograms, the system provides clinicians with a clear visual record for retrospective review. This study confirms that auditory-inspired modeling holds promise for future diagnostic software development. The evidence supports the integration of these refined techniques into routine respiratory health screening protocols.

The system processes auditory data to extract the mean frequency and duration of each sound. This quantitative information serves as the basis for the algorithm's decision-making, ensuring that only sounds meeting specific criteria are classified as wheezes.

The researchers measured the algorithm's performance across 36 subjects, including 11 individuals who exhibited wheezing. This measurement confirmed that the new approach achieves higher accuracy compared to earlier, less reliable detection techniques.

The authors suggest that this technology allows for the quantification of wheeze severity. They propose that this capability provides a significant advantage over human auscultation, which is often limited by the clinician's subjective experience and lack of a historical record.