Parkinson's Disease Diagnosis through Speech

Parkinson's disease (PD) is primarily diagnosed by clinical examinations, such as walking test, handwriting test, and MRI diagnostic. In our research, we propose to diagnose Parkinson’s disease through Speech. Classification of PD using speech records is a challenging task owing to the fact that the classification accuracy is still lower than doctor-level. Here we demonstrate classification of Parkinson’s disease using Neural Network by extracting speech features such as MFCC, HFCC, LPC, Pitch, Energy and ZCR. Mel frequency cepstral coefficients (MFCC) are the most widely used speech features in automatic speech recognition systems. Human factor cepstral coefficients (HFCC), a modification of MFCC that uses the known relationship between center frequency and critical bandwidth to decouple filter bandwidth from filter spacing had recently introduced by the researchers. In our project, we extracted the features from speech signal and then mean of that extracted signal is taken. To make a dataset of extracted signal mean is taken, then this dataset is fed to Neural Network which classify the signal as pathological or normal.