|
Descripción
|
|
|---|---|
| This study considers phonation and articulation measures to model voice disorders produced by three different pathologies: Laryngeal pathologies (LP), cleft lip and palate (CLP), and Parkinson?s disease (PD). Different speech tasks are considered including sustained vowel phonations, isolated words, and read texts. The obtained accuracies, in terms of the area under the ROC curve (AUC), range between 55.7 and 99.2 depending on the pathology. The results suggest that phonation features are appropriate to model LP; however, for the case of CLP and PD, it seems like the articulation measures provide more information about the problems in moving and controlling the articulators of the vocal tract. This work is a step towards the development of methodologies for the automatic discrimination among different voice disorders. | |
|
Internacional
|
Si |
|
Nombre congreso
|
International Work-Conference on the Interplay Between Natural and Artificial Computation, IWINAC 2017 |
|
Tipo de participación
|
960 |
|
Lugar del congreso
|
Coruña |
|
Revisores
|
Si |
|
ISBN o ISSN
|
0302-9743 |
|
DOI
|
10.1007/978-3-319-59773-7 43 |
|
Fecha inicio congreso
|
19/06/2017 |
|
Fecha fin congreso
|
23/06/2017 |
|
Desde la página
|
424 |
|
Hasta la página
|
434 |
|
Título de las actas
|
Natural and Artificial Computation for Biomedicine and Neuroscience |