Descripción
|
|
---|---|
There has been much recent research into the connection between Parkinson's disease (PD) and speech impairment. The findings prove the suitability of monitoring voice features, for tracking PD severity. The Patient Voice Analysis Project (https://www.synapse.org) is a public on line platform in which PD patients provide voice recordings and information about their symptoms. This platform have great potential to obtain large databases that, in addition to facilitate patient monitoring, help in PD research. The PVA data set consist of 779 remote, non-invasive, self administrated speech test, with sustained vowel phonation, and self reported outcomes of Parkinson's Disease Rating Scale questionnaire. With this data set, we have used machine supervised learning classification algorithms for modeling PD severity (five classes) from voice recordings of patients. The best results for the multi-class classification problem have been obtained with a KNN algorithm and features selection. Results show how public telemonitoring platforms could facilitate large-scale clinical trials into novel PD treatments. | |
Internacional
|
Si |
Nombre congreso
|
Advances in Data Mining 17th Industrial Conference on Data Mining. |
Tipo de participación
|
970 |
Lugar del congreso
|
New-York, |
Revisores
|
Si |
ISBN o ISSN
|
978-39-4295-250-7 |
DOI
|
|
Fecha inicio congreso
|
12/07/2017 |
Fecha fin congreso
|
16/07/2017 |
Desde la página
|
121 |
Hasta la página
|
131 |
Título de las actas
|
Advances in Data Mining 17th Industrial Conference on Data Mining. Publicado por Ibai-Publishing, ISSN:1864-9734 |