|
Descripción
|
|
|---|---|
| Neurological diseases (ND) are a growing concern nowadays in developed societies due to the growth of life expectancy. The tasks of diagnosing, treating, monitoring and assisting patients with ND is demanding more resources in a world with difficult economic and social challenges. It is known since long time ago that voice and speech are vehicular conveyors of psychological and neurological disease, and as such they have been used in subjective evaluation of ND. The present paper aim is two-fold: in one hand it reviews voice and speech correlates which may be used by automatic machine learning applications in producing objective evaluation of ND to comply with the main tasks demanded; on the other hand a new methodology based on phonation biomechanical correlates is presented to offer a description of Parkinson's Disease (PD) voice correlates which may be used in grading and monitoring patients. The paper presents the main characteristics of the methodology and provides results from a set of PD and control study cases. The first important result is that all cases studied present tremor in voice, which is very low in amplitude in control cases, but significantly large in PD cases. Cases labelled subjectively as tremor-free would coincide with tremor frequencies over 8-10 Hz. In all PD cases tremor relative amplitude and vocal-fold stiffness statistical dispersion were well above average estimates for the control population. Results include many other indicators which could be used to implement applications for objective evaluation and monitoring of PD and other related diseases. | |
|
Internacional
|
Si |
|
ISSN o ISBN
|
978-989-8565-36-5 |
|
Entidad relacionada
|
INSTIC |
|
Nacionalidad Entidad
|
PORTUGAL |
|
Lugar del congreso
|
Barcelona, España |