EUVIMEDEUVIMED entdecken
Crossref

An Intelligent Low-Cost Digital Stethoscope Using Arduino, Signal Processing, and Machine Learning for Real-Time Heart Sound Analysis

Nafisa M. Mapari, Alefiya J. Managori, Hajra A. Naik, Monika G. Chaurasiya, Tasmiya I. Kandhal

Journal of Smart Sensors and Computing · 03.06.2025 · 2 (2)

DOI: 10.64189/ssc.26206 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Smart Sensors and Computing, online erschienen am 03.06.2025. Verfasst von Nafisa M. Mapari und 4 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Cardiovascular diseases remain among the leading causes of mortality worldwide, creating a growing need for affordable and accessible cardiac monitoring systems. Traditional acoustic stethoscopes rely heavily on physician expertise and do not provide capabilities such as signal visualization, storage, or automated analysis. To address these limitations, this paper presents an intelligent low-cost digital stethoscope system using Arduino, signal processing, and machine learning for real-time heart sound analysis. The proposed system uses an Arduino UNO microcontroller integrated with a modified KY-037 heart sound acquisition module and an external electret microphone coupled with a stethoscope chest piece for phonocardiogram acquisition. The acquired signals are transmitted through USB serial communication and processed using a multistage signal preprocessing pipeline consisting of median filtering, Butterworth bandpass filtering, moving average smoothing, and outlier suppression to improve signal quality and reduce environmental noise. The processed heart sound signals are analyzed for real-time heart rate estimation using adaptive peak detection techniques. In addition, a machine learning-based cardiac sound classification module is implemented using a random forest classifier trained on the PhysioNet CirCor DigiScope phonocardiogram dataset. Feature extraction is performed using Mel Frequency Cepstral Coefficients (MFCCs), spectral features, chroma features, and signal energy parameters. The developed system also includes a Streamlit-based monitoring dashboard for live waveform visualization, BPM tracking, patient record management, data storage, CSV export, and PDF report generation. Experimental evaluation demonstrated successful real-time heart sound acquisition and an overall classification accuracy of 82%, with effective waveform visualization and stable BPM estimation. The proposed system provides a compact, economical, and scalable solution for intelligent cardiac monitoring and has potential applications in low-resource healthcare environments, telemedicine, and biomedical education.

Zitieren

APA

Mapari, N. M., Managori, A. J., Naik, H. A., Chaurasiya, M. G., & Kandhal, T. I. (2025). An Intelligent Low-Cost Digital Stethoscope Using Arduino, Signal Processing, and Machine Learning for Real-Time Heart Sound Analysis. Journal of Smart Sensors and Computing, 2(2). https://doi.org/10.64189/ssc.26206

Vancouver

Mapari NM, Managori AJ, Naik HA, Chaurasiya MG, Kandhal TI. An Intelligent Low-Cost Digital Stethoscope Using Arduino, Signal Processing, and Machine Learning for Real-Time Heart Sound Analysis. Journal of Smart Sensors and Computing. 2025;2(2). doi:10.64189/ssc.26206

BibTeX
@article{Mapari2025,
  author = {Mapari, Nafisa M. and Managori, Alefiya J. and Naik, Hajra A. and Chaurasiya, Monika G. and Kandhal, Tasmiya I.},
  title = {An Intelligent Low-Cost Digital Stethoscope Using Arduino, Signal Processing, and Machine Learning for Real-Time Heart Sound Analysis},
  journal = {Journal of Smart Sensors and Computing},
  year = {2025},
  volume = {2},
  number = {2},
  doi = {10.64189/ssc.26206}
}
Kostenlos starten →