Towards Device-Agnostic Mobile Cough Detection with Convolutional Neural Networks
Type
conference paper
Date Issued
2019-06
Author(s)
Abstract
Ubiquitous mobile devices have the potential to reduce the financial burden of healthcare systems by providing scalable and cost-efficient health monitoring applications. Coughing is a symptom associated with prevalent pulmonary diseases, and bears great potential for being exploited by monitoring applications. Prior research has shown the feasibility of cough detection by smartphone-based audio recordings, but it is still open as to whether current detection models generalize well to a variety of mobile devices to ensure scalability. We first conducted a lab study with 43 subjects and recorded 6737 cough samples and 8854 control sounds by 5 different recording devices. We then reimplemented two approaches from prior work and investigated their performance in two different scenarios across devices. We propose an efficient convolutional neural network architecture and an ensemble based classifier to reduce the cross-device discrepancy. Our approach produced mean accuracies in the range [85.9%, 90.9%], showing consistency across devices (SD = [1.5%, 2.7%]) and outperforming prior learning algorithms. Thus, our proposal is a step towards cost-efficient, ubiquitous, scalable and device-agnostic cough detection.
Language
English
HSG Classification
contribution to scientific community
HSG Profile Area
SoM - Business Innovation
Book title
2019 IEEE International Conference on Healthcare Informatics (ICHI)
Publisher
IEEE
Publisher place
Xi’an, China
Event Title
2019 IEEE International Conference on Healthcare Informatics (ICHI)
Event Location
Xi’an, China
Event Date
June 10-13, 2019
Division(s)
Eprints ID
258043
File(s)![Thumbnail Image]()
Loading...
open.access
Name
Barata et al 2019 Device-Agnostic Mobile Cough Detection.pdf
Size
1.36 MB
Format
Adobe PDF
Checksum (MD5)
8dfebac4be577dfbaf8813cbf59b0db5