TY - JOUR
T1 - COVID-19 detection using cough sound analysis and deep learning algorithms
AU - Rao, Sunil
AU - Narayanaswamy, Vivek
AU - Esposito, Michael
AU - Thiagarajan, Jayaraman J.
AU - Spanias, Andreas
N1 - Funding Information:
This work was funded by the REU supplement NSF NCSS/SenSIP I/UCRC Award number 1540040. We would also like to thank the authors in [57] for their insightful suggestions on model development.
Publisher Copyright:
© 2021 - IOS Press. All rights reserved.
PY - 2021
Y1 - 2021
N2 - Reliable and rapid non-invasive testing has become essential for COVID-19 diagnosis and tracking statistics. Recent studies motivate the use of modern machine learning (ML) and deep learning (DL) tools that utilize features of coughing sounds for COVID-19 diagnosis. In this paper, we describe system designs that we developed for COVID-19 cough detection with the long-term objective of embedding them in a testing device. More specifically, we use log-mel spectrogram features extracted from the coughing audio signal and design a series of customized deep learning algorithms to develop fast and automated diagnosis tools for COVID-19 detection. We first explore the use of a deep neural network with fully connected layers. Additionally, we investigate prospects of efficient implementation by examining the impact on the detection performance by pruning the fully connected neural network based on the Lottery Ticket Hypothesis (LTH) optimization process. In general, pruned neural networks have been shown to provide similar performance gains to that of unpruned networks with reduced computational complexity in a variety of signal processing applications. Finally, we investigate the use of convolutional neural network architectures and in particular the VGG-13 architecture which we tune specifically for this application. Our results show that a unique ensembling of the VGG-13 architecture trained using a combination of binary cross entropy and focal losses with data augmentation significantly outperforms the fully connected networks and other recently proposed baselines on the DiCOVA 2021 COVID-19 cough audio dataset. Our customized VGG-13 model achieves an average validation AUROC of 82.23% and a test AUROC of 78.3% at a sensitivity of 80.49%.
AB - Reliable and rapid non-invasive testing has become essential for COVID-19 diagnosis and tracking statistics. Recent studies motivate the use of modern machine learning (ML) and deep learning (DL) tools that utilize features of coughing sounds for COVID-19 diagnosis. In this paper, we describe system designs that we developed for COVID-19 cough detection with the long-term objective of embedding them in a testing device. More specifically, we use log-mel spectrogram features extracted from the coughing audio signal and design a series of customized deep learning algorithms to develop fast and automated diagnosis tools for COVID-19 detection. We first explore the use of a deep neural network with fully connected layers. Additionally, we investigate prospects of efficient implementation by examining the impact on the detection performance by pruning the fully connected neural network based on the Lottery Ticket Hypothesis (LTH) optimization process. In general, pruned neural networks have been shown to provide similar performance gains to that of unpruned networks with reduced computational complexity in a variety of signal processing applications. Finally, we investigate the use of convolutional neural network architectures and in particular the VGG-13 architecture which we tune specifically for this application. Our results show that a unique ensembling of the VGG-13 architecture trained using a combination of binary cross entropy and focal losses with data augmentation significantly outperforms the fully connected networks and other recently proposed baselines on the DiCOVA 2021 COVID-19 cough audio dataset. Our customized VGG-13 model achieves an average validation AUROC of 82.23% and a test AUROC of 78.3% at a sensitivity of 80.49%.
KW - COVID-19 detection
KW - artificial intelligence
KW - deep learning
KW - healthcare
KW - machine learning
KW - respiratory diagnosis
UR - http://www.scopus.com/inward/record.url?scp=85123957012&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85123957012&partnerID=8YFLogxK
U2 - 10.3233/IDT-210206
DO - 10.3233/IDT-210206
M3 - Article
AN - SCOPUS:85123957012
SN - 1872-4981
VL - 15
SP - 655
EP - 665
JO - Intelligent Decision Technologies
JF - Intelligent Decision Technologies
IS - 4
ER -