Wireless Mobile Communication and Healthcare. Third International Conference, MobiHealth 2012, Paris, France, November 21-23, 2012, Revised Selected Papers

Research Article

A Novel Method for Feature Extraction in Vocal Fold Pathology Diagnosis

Download89 downloads
  • @INPROCEEDINGS{10.1007/978-3-642-37893-5_11,
        author={Vahid Majidnezhad and Igor Kheidorov},
        title={A Novel Method for Feature Extraction in Vocal Fold Pathology Diagnosis},
        proceedings={Wireless Mobile Communication and Healthcare. Third International Conference, MobiHealth 2012, Paris, France, November 21-23, 2012, Revised Selected Papers},
        proceedings_a={MOBIHEALTH},
        year={2013},
        month={4},
        keywords={Vocal fold pathology diagnosis Wavelet Packet Decomposition Mel-Frequency-Cepstral-Coefficients (MFCCs) Energy Entropy Support Vector Machine (SVM)},
        doi={10.1007/978-3-642-37893-5_11}
    }
    
  • Vahid Majidnezhad
    Igor Kheidorov
    Year: 2013
    A Novel Method for Feature Extraction in Vocal Fold Pathology Diagnosis
    MOBIHEALTH
    Springer
    DOI: 10.1007/978-3-642-37893-5_11
Vahid Majidnezhad1,*, Igor Kheidorov1,*
  • 1: Islamic Azad University
*Contact email: vahidmn@yahoo.com, ikheidorov@sakrament.com

Abstract

Acoustic analysis is a proper method in vocal fold pathology diagnosis so that it can complement and in some cases replace the other invasive, based on direct vocal fold observation, methods. There are different approaches for vocal fold pathology diagnosis. These algorithms usually have two stages which are Feature Extraction and Classification. While the second stage implies a choice of a variety of machine learning methods, the first stage plays a critical role in performance of the classification system. In this paper, three types of features which are Energy and Entropy resulting from the Wavelet Packet Tree and Mel-Frequency-Cepstral-Coefficients (MFCCs), and also their combination are investigated. Finally a new type of feature vector, based on Energy and Mel-Frequency-Cepstral-Coefficients, is proposed. Support vector machine is used as a classifier for evaluating the performance of our proposed method. The results show the priority of the proposed method in comparison with other methods.