Diagnosis of voice disorders using band energy spectrum in wavelet domain

Abstract

In the evolution of quality of speech, acoustic analyses of normal and pathological voices have become increasingly interesting to researchers in laryngology and speech pathologies. Vocal signal information plays an important role to understand the process of vocal fold pathology formation. Noninvasive acoustical analysis of vocal signal helps to perform accurate diagnoses of voice disorders. This paper presents a feature extraction method using band energy spectrum and wavelet analysis for the diagnosis of vocal pathology. Two simple neural network models are developed for testing the feature extraction procedure and for the classification of pathological voices. The simulation results indicate that the proposed algorithms distinguish the voice as pathological or a non-pathological voice accurately and the proposed method can be used as diagnosis tool for the detection of voice disorders.