Descripción
|
|
---|---|
Parkinson's disease is manifested as well in handwriting as in voice. Previous researches have carried out different procedures to estimate the dysfunctions of the illness in voice and handwriting separately. This paper proposes one parameter to evaluate the influence of the illness on both voice and handwriting as the symptoms affecting both has a common origin. Specifically, the parameter proposed is based on the Kinematic Theory of rapid human movements. It allows to quantify the deficits caused by Parkinson?s disease in both handwriting and voice. The velocity profile obtained to characterize voice between the first and second formant is computed by a spatio-temporal approximation. In handwriting, the velocity profile is obtained from the sampled positions of the pen on a digital tablet. Once the velocity profile is derived, it is transformed to fit into the lognormal model in which similarities between voice and handwriting has been found for performance of these tasks by Parkinson?s patients. The experiments with different databases of voice and handwriting recorded from different patients in different labs display encouraging results. | |
Internacional
|
Si |
Nombre congreso
|
Int. Conf. on Pattern Recognition and Artificial Intelligence (ICPRAI 2018) |
Tipo de participación
|
960 |
Lugar del congreso
|
Montreal, Canada |
Revisores
|
Si |
ISBN o ISSN
|
1-895193-06-0 |
DOI
|
|
Fecha inicio congreso
|
13/05/2018 |
Fecha fin congreso
|
17/05/2018 |
Desde la página
|
795 |
Hasta la página
|
799 |
Título de las actas
|
Proc. of the Int. Conf. on Pattern Recognition and Artificial Intelligence (ICPRAI 2018) |