The presented research positively evaluates vocality of the cough sound by estimating the averaged cough fundamental frequency or pitch. The fundamental frequency was determined by autocorrelation analysis on both the rough time-signal and the linear predicted time-signal. The experimental cough database was registered in the free acoustical field on respectively 3 pathological and 9 healthy non-smoking human subjects and on 2 pathological and 2 healthy Belgian Landrace piglets. For both species differences between pitch values for cough-sounds originating from subjects suffering from a respiratory infection and healthy subjects are put forward. The retrieved pitch-differences between respectively healthy and infected subjects indicate the existence of acoustical different cough-classes in accordance with a different cause or physical condition of the respiratory system.
Index Terms. Physiological signal processing; Sound analysis; Cough; Biomedical computing; Pitch extraction
Cite as: Hirtum, A.v., Berckmans, D. (2001) Assessing the sound of cough towards vocality. Proc. Models and Analysis of Vocal Emissions for Biomedical Applications (MAVEBA 2001), 222-227
@inproceedings{hirtum01_maveba,
author={Annemie van Hirtum and Daniel Berckmans},
title={{Assessing the sound of cough towards vocality}},
year=2001,
booktitle={Proc. Models and Analysis of Vocal Emissions for Biomedical Applications (MAVEBA 2001)},
pages={222--227}
}