Audio Feature Ranking for Sound-Based COVID-19 Patient Detection

Julia A. Meister; Khuong An Nguyen; Zhiyuan Luo

Audio Feature Ranking for Sound-Based COVID-19 Patient Detection

Julia A. Meister, Khuong An Nguyen, Zhiyuan Luo

Research output: Chapter in Book/Conference proceeding with ISSN or ISBN › Conference contribution with ISSN or ISBN › peer-review

Abstract

Audio classification using breath and cough samples has recently emerged as a low-cost, non-invasive, and accessible COVID-19 screening method. However, a comprehensive survey shows that no application has been approved for official use at the time of writing, due to the stringent reliability and accuracy requirements of the critical healthcare setting. To support the development of Machine Learning classification models, we performed an extensive comparative investigation and ranking of 15 audio features, including less well-known ones. The results were verified on two independent COVID-19 sound datasets. By using the identified top-performing features, we have increased COVID-19 classification accuracy by up to 17% on the Cambridge dataset and up to 10% on the Coswara dataset compared to the original baseline accuracies without our feature ranking.

Original language	English
Title of host publication	Progress in Artificial Intelligence
Subtitle of host publication	21st EPIA Conference on Artificial Intelligence
Editors	Goreti Marreiros, Bruno Martins, Ana Paiva, Bernardete Ribeiro, Alberto Sardinha
Publisher	Springer
Pages	146-158
Number of pages	13
ISBN (Electronic)	9783031164743
ISBN (Print)	9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164743, 9783031164736
Publication status	Published - 13 Sept 2022
Event	21st EPIA Conference on Artificial Intelligence: Progress in Artificial Intelligence - Lisbon, Portugal Duration: 31 Aug 2022 → 2 Sept 2022 https://epia2022.inesc-id.pt/

Publication series

Name	Lecture Notes in Computer Science (LNCS)
Publisher	Springer
Volume	13566

Conference

Conference	21st EPIA Conference on Artificial Intelligence
Abbreviated title	EPIA 2022
Country/Territory	Portugal
City	Lisbon
Period	31/08/22 → 2/09/22
Internet address	https://epia2022.inesc-id.pt/

Keywords

COVID-19 classification
Audio event engineering
Sound feature ranking

Cite this

Meister, Julia A. ; Nguyen, Khuong An ; Luo, Zhiyuan. / Audio Feature Ranking for Sound-Based COVID-19 Patient Detection. Progress in Artificial Intelligence : 21st EPIA Conference on Artificial Intelligence. editor / Goreti Marreiros ; Bruno Martins ; Ana Paiva ; Bernardete Ribeiro ; Alberto Sardinha. Springer, 2022. pp. 146-158 (Lecture Notes in Computer Science (LNCS)).

@inproceedings{cbae2ee9fec949cbad6d186f86197704,

title = "Audio Feature Ranking for Sound-Based COVID-19 Patient Detection",

abstract = "Audio classification using breath and cough samples has recently emerged as a low-cost, non-invasive, and accessible COVID-19 screening method. However, a comprehensive survey shows that no application has been approved for official use at the time of writing, due to the stringent reliability and accuracy requirements of the critical healthcare setting. To support the development of Machine Learning classification models, we performed an extensive comparative investigation and ranking of 15 audio features, including less well-known ones. The results were verified on two independent COVID-19 sound datasets. By using the identified top-performing features, we have increased COVID-19 classification accuracy by up to 17% on the Cambridge dataset and up to 10% on the Coswara dataset compared to the original baseline accuracies without our feature ranking.",

keywords = "COVID-19 classification, Audio event engineering, Sound feature ranking",

author = "Meister, {Julia A.} and Nguyen, {Khuong An} and Zhiyuan Luo",

year = "2022",

month = sep,

day = "13",

language = "English",

isbn = "9783031164743",

series = "Lecture Notes in Computer Science (LNCS)",

publisher = "Springer",

pages = "146--158",

editor = "Goreti Marreiros and Bruno Martins and Ana Paiva and Bernardete Ribeiro and Alberto Sardinha",

booktitle = "Progress in Artificial Intelligence",

note = "21st EPIA Conference on Artificial Intelligence : Progress in Artificial Intelligence, EPIA 2022 ; Conference date: 31-08-2022 Through 02-09-2022",

url = "https://epia2022.inesc-id.pt/",

}

Meister, JA, Nguyen, KA & Luo, Z 2022, Audio Feature Ranking for Sound-Based COVID-19 Patient Detection. in G Marreiros, B Martins, A Paiva, B Ribeiro & A Sardinha (eds), Progress in Artificial Intelligence : 21st EPIA Conference on Artificial Intelligence. Lecture Notes in Computer Science (LNCS), vol. 13566, Springer, pp. 146-158, 21st EPIA Conference on Artificial Intelligence, Lisbon, Portugal, 31/08/22.

Audio Feature Ranking for Sound-Based COVID-19 Patient Detection. / Meister, Julia A.; Nguyen, Khuong An; Luo, Zhiyuan.
Progress in Artificial Intelligence : 21st EPIA Conference on Artificial Intelligence. ed. / Goreti Marreiros; Bruno Martins; Ana Paiva; Bernardete Ribeiro; Alberto Sardinha. Springer, 2022. p. 146-158 (Lecture Notes in Computer Science (LNCS); Vol. 13566).

Research output: Chapter in Book/Conference proceeding with ISSN or ISBN › Conference contribution with ISSN or ISBN › peer-review

TY - GEN

T1 - Audio Feature Ranking for Sound-Based COVID-19 Patient Detection

AU - Meister, Julia A.

AU - Nguyen, Khuong An

AU - Luo, Zhiyuan

PY - 2022/9/13

Y1 - 2022/9/13

N2 - Audio classification using breath and cough samples has recently emerged as a low-cost, non-invasive, and accessible COVID-19 screening method. However, a comprehensive survey shows that no application has been approved for official use at the time of writing, due to the stringent reliability and accuracy requirements of the critical healthcare setting. To support the development of Machine Learning classification models, we performed an extensive comparative investigation and ranking of 15 audio features, including less well-known ones. The results were verified on two independent COVID-19 sound datasets. By using the identified top-performing features, we have increased COVID-19 classification accuracy by up to 17% on the Cambridge dataset and up to 10% on the Coswara dataset compared to the original baseline accuracies without our feature ranking.

AB - Audio classification using breath and cough samples has recently emerged as a low-cost, non-invasive, and accessible COVID-19 screening method. However, a comprehensive survey shows that no application has been approved for official use at the time of writing, due to the stringent reliability and accuracy requirements of the critical healthcare setting. To support the development of Machine Learning classification models, we performed an extensive comparative investigation and ranking of 15 audio features, including less well-known ones. The results were verified on two independent COVID-19 sound datasets. By using the identified top-performing features, we have increased COVID-19 classification accuracy by up to 17% on the Cambridge dataset and up to 10% on the Coswara dataset compared to the original baseline accuracies without our feature ranking.

KW - COVID-19 classification

KW - Audio event engineering

KW - Sound feature ranking

UR - https://link.springer.com/chapter/10.1007/978-3-031-16474-3_13

M3 - Conference contribution with ISSN or ISBN

SN - 9783031164743

SN - 9783031164736

T3 - Lecture Notes in Computer Science (LNCS)

SP - 146

EP - 158

BT - Progress in Artificial Intelligence

A2 - Marreiros, Goreti

A2 - Martins, Bruno

A2 - Paiva, Ana

A2 - Ribeiro, Bernardete

A2 - Sardinha, Alberto

PB - Springer

T2 - 21st EPIA Conference on Artificial Intelligence

Y2 - 31 August 2022 through 2 September 2022

ER -

Audio Feature Ranking for Sound-Based COVID-19 Patient Detection

Abstract

Publication series

Conference

Keywords

Other files and links

Fingerprint

Cite this