Voice pathology detection using machine learning technique

Abstract

Recent proposed researches have witnessed that voice pathology detection systems can effectively contribute to the voice disorders assessment and provide early detection of voice pathologies. These systems used machine learning techniques which are considered as very promising tools in the detection of voice pathologies. However, most proposed systems in the detection of voice disorder utilized limited database. Furthermore, low accuracy rate is still the one of the most challenging issues for these techniques. This paper presents a voice pathology detection system using Online Sequential Extreme Learning Machine (OSELM) to classify the voice signal into healthy or pathological. In this work, the voice features are extracted by using Mel-Frequency Cepstral Coefficient (MFCC). The voice samples for the vowel /a/ were collected equally from Saarbrücken voice database (SVD). The proposed method is evaluated by three widely used measurements which are accuracy, sensitivity and specificity. The obtained results show that the maximum accuracy, sensitivity and specificity are 85%, 87% and 87%, respectively. According to the experimental results, the performance of OSELM algorithm is able to differentiate healthy and pathological voices effectively.