Nonlinear Signal Processing for Voice Disorder Detection by Using Modified GP Algorithm and Surrogate Data Analysis

Abstract

Acoustic voice analysis is an effective, cheap and non-invasive tool that can be used to confirm the initial diagnosis and provides an objective determination of the impairment. The nonlinearities of the voice source mechanisms may cause the existence of chaos in human voice production. Voice pathology can cause to addition colored noise to voice wave. Added noise to a chaotic signal causes reduction of the deterministic property and therefore increases correlation dimension of signal. Surrogate data analysis can measure this deviation and give a criterion for amount of noise added to the chaotic signal. By using this criterion a threshold level is set to separate disordered voice from normal voice and 95% accuracy is achieved.