Authors:
Fernanda M. Ribeiro
1
;
Alvaro R. Pereira Jr.
1
;
Débora M. Barroso Paiva
2
;
Luciana M. Alves
3
and
Andrea G. Campos Bianchi
1
Affiliations:
1
Computing Department, Federal University of Ouro Preto, Ouro Preto, Brazil
;
2
School of Computing, Federal University of Mato do Grosso do Sul, Campo Grande, Brazil
;
3
Fonoaudiology Department, Federal University of Minas Gerais, Belo Horizonte, Brazil
Keyword(s):
Signal Processing, Features, Mel Cepstral Frequencies, Supervised Learning, Decision Making System.
Abstract:
The pathologies of the language are alterations in the reading of a text caused by traumatisms. Many people go untreated due to the lack of specific tools and the high cost of using proprietary software, however, new audio signal processing technologies can aid in the process of identifying genetic pathologies. In the past, a methodology was developed by medical specialists, which extracts characteristics from the reading of a text aloud and returns evidence of dyslexia. In this work, a new computational approach is described in order to automate serving as a tool for dyslexia indication efficiently. The analysis is done in recordings of the reading of pre-defined texts with school-age children, being extracted characteristics using specific methodologies. The indication of the probability of dyslexia is performed using a machine learning algorithm. The tests were performed comparing with the classification performed by the specialist, obtaining high accuracy on the evidence of dysle
xia. The difference between the values of the automatically collected characteristics and the manually assigned was below 20% for most of the characteristics. Finally, the results show a very promising area for audio signal processing with respect to the aid to specialists in the decision making related to language pathologies.
(More)