Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

VoxMed: One-Step Respiratory Disease Classifier using Digital Stethoscope Sounds

Paridhi Mundra, Manik Sharma, Yashwardhan Chaudhuri, Orchid Chetia Phukan, Arun Balaji Buduru

TL;DR

Respiratory diseases demand fast, reliable bedside diagnostics. VoxMed offers a UI-assisted, one-step classifier that analyzes digital stethoscope recordings using an Audio Spectrogram Transformer (AST) for feature extraction and a 1-D CNN for disease classification. Evaluated on the ICBHI dataset, AST-based embeddings show competitive accuracy and F1 scores, outperforming several alternative backbones across multiple class configurations. The system supports on-the-spot assessment via a simple upload-and-infer workflow and can be enhanced with API-sourced patient information to aid clinical decisions. Together, VoxMed provides a practical, rapid diagnostic aid with potential to improve patient care and workflow in busy clinical environments.

Abstract

As respiratory illnesses become more common, it is crucial to quickly and accurately detect them to improve patient care. There is a need for improved diagnostic methods for immediate medical assessments for optimal patient outcomes. This paper introduces VoxMed, a UI-assisted one-step classifier that uses digital stethoscope recordings to diagnose respiratory diseases. It employs an Audio Spectrogram Transformer(AST) for feature extraction and a 1-D CNN-based architecture to classify respiratory diseases, offering professionals information regarding their patients respiratory health in seconds. We use the ICBHI dataset, which includes stethoscope recordings collected from patients in Greece and Portugal, to classify respiratory diseases. GitHub repository: https://github.com/Sample-User131001/VoxMed

VoxMed: One-Step Respiratory Disease Classifier using Digital Stethoscope Sounds

TL;DR

Respiratory diseases demand fast, reliable bedside diagnostics. VoxMed offers a UI-assisted, one-step classifier that analyzes digital stethoscope recordings using an Audio Spectrogram Transformer (AST) for feature extraction and a 1-D CNN for disease classification. Evaluated on the ICBHI dataset, AST-based embeddings show competitive accuracy and F1 scores, outperforming several alternative backbones across multiple class configurations. The system supports on-the-spot assessment via a simple upload-and-infer workflow and can be enhanced with API-sourced patient information to aid clinical decisions. Together, VoxMed provides a practical, rapid diagnostic aid with potential to improve patient care and workflow in busy clinical environments.

Abstract

As respiratory illnesses become more common, it is crucial to quickly and accurately detect them to improve patient care. There is a need for improved diagnostic methods for immediate medical assessments for optimal patient outcomes. This paper introduces VoxMed, a UI-assisted one-step classifier that uses digital stethoscope recordings to diagnose respiratory diseases. It employs an Audio Spectrogram Transformer(AST) for feature extraction and a 1-D CNN-based architecture to classify respiratory diseases, offering professionals information regarding their patients respiratory health in seconds. We use the ICBHI dataset, which includes stethoscope recordings collected from patients in Greece and Portugal, to classify respiratory diseases. GitHub repository: https://github.com/Sample-User131001/VoxMed
Paper Structure (4 sections, 2 figures, 1 table)

This paper contains 4 sections, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. Application Overview
  3. Evaluation
  4. Conclusion

Figures (2)

  • Figure 1: VoxMed Architecture: The architecture accepts a wav file for digital stethoscope sound and passes through an audio spectrogram transformer to extract features. Extracted features are passed through a 1-D CNN architecture as shown in the image to detect the type of respiratory disease.
  • Figure 2: VoxMed Workflow: VoxMed respiratory detection requires the input of a digital stethoscope recording from the patient. We upload the recording through the UI and click on submit to process. Finally, we see possible respiratory ailment, its symptoms, and more information gathered through scraping information using APIs.