Title
A forced gaussians based methodology for the differential evaluation of Parkinson's Disease by means of speech processing.
Abstract
•The Forced gaussian schemes allow to compare phonemes between different classes.•PD influences phonemes requiring a higher narrowing of the vocal tract.•The glass ceiling of accuracy in the detection of PD from speech is below 95%.•Cross-corpora validation is highly recommended to ensure generalization.
Year
DOI
Venue
2019
10.1016/j.bspc.2018.10.020
Biomedical Signal Processing and Control
Keywords
Field
DocType
Speech processing,Machine learning,Parkinson's Disease,Phoneme,Forced alignment
Speech processing,Parkinson's disease,Pattern recognition,Classification scheme,Speech recognition,Clinical diagnosis,Hypokinetic dysarthria,Artificial intelligence,Vocal tract,Mathematics
Journal
Volume
ISSN
Citations 
48
1746-8094
2
PageRank 
References 
Authors
0.42
9
7