Classification of the Heart Auscultation Signals

Primož Kocuvan, Drago Torkar


Listening to the internal body sounds (auscultation) is one of the oldest techniques in medicine to diagnose heart and lung diseases. The digital heart auscultation signals are obtained with digital electronic stethoscope and can be processed automatically to obtain some coarse indications about the heart or lung condition. There are many ways of how to process the auscultation signals and quite some were published in the last years. In this paper we present one possible set of methods to reach the goal of heart murmur recognition up to the level to distinguish between the pathological murmurs from the physiological ones. The special attention was devoted to signal feature selection and extraction where we used the distribution of signal power over frequencies as the key difference between the normal and the pathological murmurs. The whole procedure including the signal processing, the feature extraction and the comparison of four machine learning classification methods is adequately described. It was tested on a balanced and on an unbalanced dataset with the best achieved classification accuracy of 87.5%.


  1. Ahlström, C., 2006. Processing of the Phonocardiographic Signal - Methods for the Intelligent Stethoscope. Doctoral Thesis, Linkoping University, Institute of techonology, Available at: pdf.
  2. Atbi, A., Debbal, S. M., 2013. Segmentation of Pathological Signals Phonocardiogram by Using the Shannon Energy Envelogram. AJCM, 2(1,2), 1-14.
  3. Bentley, P., Nordehn, G., Coimbra, M., Mannor, S., 2011. The PASCAL Classifying Heart Sounds Challenge 2011 (CHSC2011). Available at: ml.
  4. Haney, I., Ipp, M., Feldman, W., McCrindle, B.W., 1999. Accuracy of clinical assessment of heart murmurs by office based (general practice) paediatricians, Archives of Desease in Childhood, 81, 409-412.
  5. Hedayioglu, F. L., Coimbra, M.T., Mattos S.S., 2009. A survey of audio processing algorithms for digital Stethoscopes. HEALTHINF, INSTICC Press, 425-429.
  6. Kim, C., Stern, R. M., 2012. Power-Normalized Cepstral Coefficients (PNCC) for robust speech recognition. Proceedings of the 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Kyoto, 4101-4104.
  7. Liang, H., Lukkarinen, S., Hartimo, I., 1997. Heart Sound Segmentation Algorithm Based on Heart Sound Envelolgram, Computers in Cardiology 1997, 7 - 10, Available at: ng97Heart.pdf.
  8. Shradhanjali, A., Chowdhury, S., Kumar, N., 2013. Power Spectral Density Estimation of EMG Signals Using Parametric and Non-Parametric Approach, Global Advanced Research Journal of Engineering, Technology and Innovation, 2(4), 111-117.
  9. Walker, H. K., Hall, W. D., Hurst, J. W., 1990. Clinical Methods, The History, Physical, and Laboratory Examinations. Butterworths, 3rd edition, Emory University School of Medicine, Atlanta, Georgia, Boston.
  10. Zhong L., Wan, J.,Huang, Z.,Cao, G., Xiao, B., 2013. Heart Murmur Recognition Based on Hidden Markov Model. Journal of Signal and Information Processing, 4, 140-144.

Paper Citation

in Harvard Style

Kocuvan P. and Torkar D. (2015). Classification of the Heart Auscultation Signals . In Proceedings of the International Conference on Health Informatics - Volume 1: HEALTHINF, (BIOSTEC 2015) ISBN 978-989-758-068-0, pages 534-539. DOI: 10.5220/0005264005340539

in Bibtex Style

author={Primož Kocuvan and Drago Torkar},
title={Classification of the Heart Auscultation Signals},
booktitle={Proceedings of the International Conference on Health Informatics - Volume 1: HEALTHINF, (BIOSTEC 2015)},

in EndNote Style

JO - Proceedings of the International Conference on Health Informatics - Volume 1: HEALTHINF, (BIOSTEC 2015)
TI - Classification of the Heart Auscultation Signals
SN - 978-989-758-068-0
AU - Kocuvan P.
AU - Torkar D.
PY - 2015
SP - 534
EP - 539
DO - 10.5220/0005264005340539