AI classification of respiratory illness through vocal biomarkers and a bespoke articulatory speech protocol

Tim Bashford*, Hok Shing Lau, Mark Huntly, Nathan Morgan, Adesua Iyenoma, Tom Powell, Biao Zeng

*Corresponding author for this work

Research output: Contribution to journalConference articlepeer-review

4 Downloads (Pure)

Abstract

Speech biomarkers represent a powerful indicator for detecting, monitoring and categorising neurological, psychological , pathological and pulmonary conditions. Facilitated by advances in computational power and artificial intelligence (AI) techniques, we present a novel ecosystem for data acquisition , analysis and storage, using an articulatory speech task. By automatically segmenting, aligning and extracting features from the vocal recordings, we present a feature extraction pipeline toward the classification of pathological conditions, specifically respiratory disease through recorded voice. Data is stored within a Trusted Research Environment, for which this work also presents a range of ethical considerations.
Original languageEnglish
Article number13
Number of pages5
JournalInternational Journal of Simulation, Systems, Science and Technology
Volume25
Issue number1
Publication statusPublished - Mar 2024
EventUKSim-AMSS: International Conference on Mathematical/Analytical Modelling and Computer Simulation - Emmanuel College, Cambridge, United Kingdom
Duration: 26 Mar 202428 Mar 2024
Conference number: 26th

Keywords

  • Speech biomarkers
  • respiratory disease
  • artificial intelligence
  • speech breathing

Fingerprint

Dive into the research topics of 'AI classification of respiratory illness through vocal biomarkers and a bespoke articulatory speech protocol'. Together they form a unique fingerprint.

Cite this