Title :
Normal versus pathology voice-an analysis
Author :
Raju, N. ; Priya, T. Lakshmi ; Mathini, S. ; Preethi, P.
Author_Institution :
Dept. of ECE, SASTRA Univ., Thanjavur, India
Abstract :
This paper proposes a novel method for diagnosis of dysathria using acoustic voice analysis. It is an effective and on-invasive method that could be used for confirmation of initial diagnosis and it would provide an objective determination of the vocal impairment. The acoustic parameters pitch and formants are extracted which gives differentiation of normal and pathological voices. Support Vector Machine is used to classify normal and disordered voice, the proposed algorithm offers least computational complexity with an accuracy of 80 to 90%.
Keywords :
bioacoustics; diseases; medical signal detection; medical signal processing; patient diagnosis; speech processing; support vector machines; acoustic voice analysis; dysathria diagnosis; on-invasive method; pathology voice-an analysis; pitch detection method; signal processing; support vector machine; vocal impairment; Acoustics; Databases; Diseases; Feature extraction; Pathology; Speech; Support vector machines; Support Vector Machine; acoustic parameter; non-invasie; pathology;
Conference_Titel :
Computing, Communication and Applications (ICCCA), 2012 International Conference on
Conference_Location :
Dindigul, Tamilnadu
Print_ISBN :
978-1-4673-0270-8
DOI :
10.1109/ICCCA.2012.6179188
