Application of Mel Cepstral processing and Support Vector Machines for diagnosing vocal disorders from voice recordings

The aim of this study was to assess the applicability of Mel Frequency Cepstral Coefficients (MFCC) of voice samples in diagnosing vocal nodules and polyps. Data acquisition involved recordings of sustained vowels and standardized sentences. Voice samples were analysed acoustically with the measurement of MFCC and the first three formants. Classification of Mel coefficients was performed by applying the Sammon Mapping and Support Vector Machines (SVM). For the tests conducted on 95 patients, voice disorders were detected with accuracy reaching approx. 80%.