Parallel source separation system for heart and lung sounds

In this paper, we propose a parallel source separation system designed to extract heart and lung sounds from single-channel mixtures. The proposed system is based on a non-negative matrix factorization (NMF) approach and a clustering strategy together with a soft-masking filtering. Furthermore, we p...

Full description

Saved in:
Bibliographic Details
Published inThe Journal of supercomputing Vol. 77; no. 8; pp. 8135 - 8150
Main Authors Muñoz-Montoro, A. J., Suarez-Dou, D., Cortina, R., Canadas-Quesada, F. J., Combarro, E. F.
Format Journal Article
LanguageEnglish
Published New York Springer US 2021
Springer Nature B.V
Subjects
Acoustics
Clustering
Compilers
Computer Science
High Performance Computing in Science and Engineering - CMMSE-2019
Interpreters
Lungs
Processor Architectures
Programming Languages
Real time
Separation
Heart
Parallel computing
Single-channel
Non-negative matrix factorization (NMF)
Lung
Sound source separation
Real time
Online AccessGet full text

Cover

More Information
Summary:In this paper, we propose a parallel source separation system designed to extract heart and lung sounds from single-channel mixtures. The proposed system is based on a non-negative matrix factorization (NMF) approach and a clustering strategy together with a soft-masking filtering. Furthermore, we propose an offline and online implementation of the framework which can be applied in many real-time scenarios, such as the extraction of clinical parameters, remote auscultation and breath sound analysis. Experimental results show that it is possible to achieve fast execution times, which enable a real-time behavior, combining parallel and high-performance techniques.
ISSN:0920-8542
1573-0484
DOI:10.1007/s11227-020-03616-0