Authors:
Ruben Fernández
1
;
Jose Luis Blanco
1
;
Luis A. Hernández
1
;
Eduardo López
1
;
José Alcazar
2
and
Doroteo T. Toledano
3
Affiliations:
1
Universidad Politécnica de Madrid, Spain
;
2
Hospital Clínico Universitario Málaga, Spain
;
3
Universidad Autonoma de Madrid, Spain
Keyword(s):
Apnoea, Automatic Speaker Recognition techniques, GMM, Nasalization.
Related
Ontology
Subjects/Areas/Topics:
Acoustic Signal Processing
;
Artificial Intelligence
;
Biomedical Engineering
;
Biomedical Signal Processing
;
Data Manipulation
;
Health Engineering and Technology Applications
;
Human-Computer Interaction
;
Methodologies and Methods
;
Neurocomputing
;
Neurotechnology, Electronics and Informatics
;
Pattern Recognition
;
Physiological Computing Systems
;
Sensor Networks
;
Soft Computing
Abstract:
The aim of this paper is to study new possibilities of using Automatic Speaker Recognition techniques (ASR) for detection of patients with severe obstructive sleep apnoea (OSA). Early detection of severe apnoea cases can be very useful to give priority to their early treatment optimizing the expensive and time-consuming tests of current diagnosis methods based on full overnight sleep in a hospital. This work is part of an on-going collaborative project between medical and signal processing communities to promote new research efforts on automatic OSA diagnosis through speech processing technologies applied on a carefully designed speech database of healthy subjects and apnoea patients. So far, in this contribution we present and discuss several approaches of applying generative Gaussian Mixture Models (GMMs), generally used in ASR systems, to model specific acoustic properties of continuous speech signals in different linguistic contexts reflecting discriminative physiological charact
eristics found in OSA patients. Finally, experimental results on the discriminative power of speaker recognition techniques adapted to severe apnoea detection are presented. These results obtain a correct classification rate of 81.25%, representing a promising result underlining the interest of this research framework and opening further perspectives for improvement using more specific speech recognition technologies.
(More)