Pathological Voice Analysis to Detect Neurological Disorders Using MFCC and SVM

The neurological problems like brian tumour, lesions, neural degeneration or regeneration, brain injury may affect the speech producing centre present in brain. Thus the voice contains the hidden information about disorders of nervous system. The use of speech processing algorithms, can detect the pathological condition of the brain. This paper investigates the adaptation of Melfrequency Cepstral Coefficients (MFCC) and Support Vector Machine (SVM) for the diagnosis of neurological disorders using voice signal. The aim of this work is to provide a novel method, which is easy-to-use, fast, noninvasive for the patient, and affordable for the clinicians. The system designed for this particular task relies on the MFCCs for feature extraction and SVM for classification.