ALZHEIMER'S DISEASE DIAGNOSIS USING SPONTANEOUS SPEECH SIGNALS AND HYBRID FEATURES

The purpose of this study is to classify spontaneous speech signals in order to automatic diagnosis of Alzheimer's disease (AD) using adaptive neuro-fuzzy inference system (ANFIS). The proposed system uses three feature sets, Lyapunov exponents as nonlinear features, acoustic features and Lyapunov exponents with acoustic features, to achieve high detection accuracy. To evaluate the performance of the method, total classification accuracy is estimated. The classification results demonstrate that the Lyapunov exponents are useful parameters which contain comprehensive information about signals. They also show the Lyapunov exponents with the ANFIS have better performance than acoustic features. The proposed method is also able to diagnose the earliest stage of AD. Therefore our method can be a spontaneous speech directed test for pre-clinical evaluation of AD diagnosis.