Advertisement

Effective Speech Features for Distinguishing Mild Dementia Patients from Healthy Person

  • Kazu NishikawaEmail author
  • Rin Hirakawa
  • Hideki Kawano
  • Kenichi Nakashi
  • Yoshihisa Nakatoh
Conference paper
  • 302 Downloads
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 1253)

Abstract

The questionnaire method is generally used for present dementia screening. However, this method requires time for 10 to 15 min with a doctor and a clinical psychologist, which puts a burden on hospitals and test subjects. The purpose of this study is to reduce the burden of users by constructing a system to distinguish patients with mild dementia and healthy persons from speech data. Before that this paper examines the effectiveness of speech features. MFCC has been confirmed to be effective in previous research, this paper extracted six kinds of other speech features that are likely to be correlated with symptoms of dementia. This paper got about 90% accuracy rate for a sentence of conversational speech in SVM and Random Forest. Moreover, this paper has calculated the importance of the features by using the SVM-RFE method. As a result, this showed that log-mel spectrum was more important than MFCC.

Keywords

Alzheimer Dementia MCI SVM Random Forest Dementia screening 

References

  1. 1.
    Toshiharu, N., Mio, O.: Japanese Perspective on Dietary Patterns and Risk of Dementia. Ministry of Health, Japanese Perspective on Dietary Patterns and Risk of Dementia. Academic Press, Oxford, pp. 285–294 (2014)Google Scholar
  2. 2.
    Tsukasa, K.: Preparation of the revised hasegawa’s simplified intelligence scale (HDS-R). Jpn J. Geriatric Psychiatry 1339–1347 (1991)Google Scholar
  3. 3.
    König, A., Satt, A., Sorin, A., et al.: Automatic speech analysis for the assessment of patients with predementia and Alzheimer’s disease. Alzheimer’s Dement. Diagn. Assess. Dis. Monit. 1, 112–124 (2015)Google Scholar
  4. 4.
  5. 5.
    Bjorn, S., Stefan, S., Anton, B.: The INTERSPEECH 2009 emotion challenge. In: ISCA, 6–10 September, Brighton UK, pp. 312–314 (2009)Google Scholar
  6. 6.
    Bjorn, S., Stefan, S., Anton, B., Felix, B., Laurence, D., Christian, M., Shrikanth, N.: The INTERSPEECH 2010 paralinguistic challenge. In: ISCA, 26–30 September, Makuhari, Chiba, Japan, pp. 2794–2797 (2010)Google Scholar
  7. 7.
    Yuki, K.: Analysis of dementia tendency of the elderly using acoustic features. Summary of graduation thesis from Department of Information Science, Aichi Prefectural University (2018)Google Scholar
  8. 8.
    Florian, E., Felix, W., Martin, W., Bj¨orn, S.: openSMILE the Munich open Speech and Music Interpretation by Large Space Extraction toolkit, p. 78 (2018)Google Scholar
  9. 9.
    Sadaoki, F.: Sound and Speech Engineering, pp. 181–184. Kindai Kagaku Sha Co. Ltd., Tokyo (1992)Google Scholar
  10. 10.
    Tokyo Medical Association: Caregiver and Community Care Guidebook, pp. 171–172 (2011)Google Scholar
  11. 11.
    Toshiya, F., Eiyai, L., Soutaro, H.: Frontotemporal dementia presenting initially with foreign accent syndrome. A novel clinical sign? pp. 397–407 (2006)Google Scholar
  12. 12.
    Vladimir, N.: V: Statistical Learning Theory. Wiley, New York (1998)Google Scholar
  13. 13.
    Leo, B.: Random Forests. Mach. Learn. 45, 5–32 (2001)zbMATHCrossRefGoogle Scholar
  14. 14.
    Daisaku, K., Kaoru, I., Shoko, W.: Detecting early stage dementia based on natural language processing. Jpn. Soc. Artif. Intell. 34(4), 1–7 (2019)Google Scholar
  15. 15.
  16. 16.
  17. 17.
    Guyon, I., Weston, J., Barnhill, S., Vapnik, V.: Gene selection for cancer classification using support vector machines. Mach. Learn. 46(1–3), 389–422 (2002)zbMATHCrossRefGoogle Scholar
  18. 18.
    Hendrik, P., Bo, L., Tuomas, V.: Deep learning for audio signal processing. J. Sel. Top. Sig. Process. 13(2), 206–219 (2019)CrossRefGoogle Scholar

Copyright information

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021

Authors and Affiliations

  • Kazu Nishikawa
    • 1
    Email author
  • Rin Hirakawa
    • 1
  • Hideki Kawano
    • 1
  • Kenichi Nakashi
    • 1
  • Yoshihisa Nakatoh
    • 1
  1. 1.Kyushu Institute of TechnologyKitakyushu-shiJapan

Personalised recommendations