Cargando…
Covid-19 recognition from cough sounds using lightweight separable-quadratic convolutional network
Automatic classification of cough data can play a vital role in early detection of Covid-19. Lots of Covid-19 symptoms are somehow related to the human respiratory system, which affect sound production organs. As a result, anomalies in cough sound is expected to be discovered in Covid-19 patients as...
Autores principales: | , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
Elsevier Ltd.
2022
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8590951/ https://www.ncbi.nlm.nih.gov/pubmed/34804190 http://dx.doi.org/10.1016/j.bspc.2021.103333 |
_version_ | 1784599102496964608 |
---|---|
author | Soltanian, Mohammad Borna, Keivan |
author_facet | Soltanian, Mohammad Borna, Keivan |
author_sort | Soltanian, Mohammad |
collection | PubMed |
description | Automatic classification of cough data can play a vital role in early detection of Covid-19. Lots of Covid-19 symptoms are somehow related to the human respiratory system, which affect sound production organs. As a result, anomalies in cough sound is expected to be discovered in Covid-19 patients as a sign of infection. This drives the research towards detection of potential Covid-19 cases with inspecting cough sound. While there are several well-performing deep networks, which are capable of classifying sound with a high accuracy, they are not suitable for using in early detection of Covid-19 as they are huge and power/memory hungry. Actually, cough recognition algorithms need to be implemented in hand-held or wearable devices in order to generate early Covid-19 warning without the need to refer individuals to health centers. Therefore, accurate and at the same time lightweight classifiers are needed, in practice. So, there is a need to either compress the complicated models or design light-weight models from the beginning which are suitable for implementation on embedded devices. In this paper, we follow the second approach. We investigate a new lightweight deep learning model to distinguish Covid and Non-Covid cough data. This model not only achieves the state of the art on the well-known and publicly available Virufy dataset, but also is shown to be a good candidate for implementation in low-power devices suitable for hand-held applications. |
format | Online Article Text |
id | pubmed-8590951 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2022 |
publisher | Elsevier Ltd. |
record_format | MEDLINE/PubMed |
spelling | pubmed-85909512021-11-15 Covid-19 recognition from cough sounds using lightweight separable-quadratic convolutional network Soltanian, Mohammad Borna, Keivan Biomed Signal Process Control Article Automatic classification of cough data can play a vital role in early detection of Covid-19. Lots of Covid-19 symptoms are somehow related to the human respiratory system, which affect sound production organs. As a result, anomalies in cough sound is expected to be discovered in Covid-19 patients as a sign of infection. This drives the research towards detection of potential Covid-19 cases with inspecting cough sound. While there are several well-performing deep networks, which are capable of classifying sound with a high accuracy, they are not suitable for using in early detection of Covid-19 as they are huge and power/memory hungry. Actually, cough recognition algorithms need to be implemented in hand-held or wearable devices in order to generate early Covid-19 warning without the need to refer individuals to health centers. Therefore, accurate and at the same time lightweight classifiers are needed, in practice. So, there is a need to either compress the complicated models or design light-weight models from the beginning which are suitable for implementation on embedded devices. In this paper, we follow the second approach. We investigate a new lightweight deep learning model to distinguish Covid and Non-Covid cough data. This model not only achieves the state of the art on the well-known and publicly available Virufy dataset, but also is shown to be a good candidate for implementation in low-power devices suitable for hand-held applications. Elsevier Ltd. 2022-02 2021-11-15 /pmc/articles/PMC8590951/ /pubmed/34804190 http://dx.doi.org/10.1016/j.bspc.2021.103333 Text en © 2021 Elsevier Ltd. All rights reserved. Since January 2020 Elsevier has created a COVID-19 resource centre with free information in English and Mandarin on the novel coronavirus COVID-19. The COVID-19 resource centre is hosted on Elsevier Connect, the company's public news and information website. Elsevier hereby grants permission to make all its COVID-19-related research that is available on the COVID-19 resource centre - including this research content - immediately available in PubMed Central and other publicly funded repositories, such as the WHO COVID database with rights for unrestricted research re-use and analyses in any form or by any means with acknowledgement of the original source. These permissions are granted for free by Elsevier for as long as the COVID-19 resource centre remains active. |
spellingShingle | Article Soltanian, Mohammad Borna, Keivan Covid-19 recognition from cough sounds using lightweight separable-quadratic convolutional network |
title | Covid-19 recognition from cough sounds using lightweight separable-quadratic convolutional network |
title_full | Covid-19 recognition from cough sounds using lightweight separable-quadratic convolutional network |
title_fullStr | Covid-19 recognition from cough sounds using lightweight separable-quadratic convolutional network |
title_full_unstemmed | Covid-19 recognition from cough sounds using lightweight separable-quadratic convolutional network |
title_short | Covid-19 recognition from cough sounds using lightweight separable-quadratic convolutional network |
title_sort | covid-19 recognition from cough sounds using lightweight separable-quadratic convolutional network |
topic | Article |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8590951/ https://www.ncbi.nlm.nih.gov/pubmed/34804190 http://dx.doi.org/10.1016/j.bspc.2021.103333 |
work_keys_str_mv | AT soltanianmohammad covid19recognitionfromcoughsoundsusinglightweightseparablequadraticconvolutionalnetwork AT bornakeivan covid19recognitionfromcoughsoundsusinglightweightseparablequadraticconvolutionalnetwork |