Abstract
Speech is a measurable behavior that can be used as a biomarker for various mental states including schizophrenia and depression. In this paper we show that simple temporal domain features, extracted from conversational speech, may highlight alterations in acoustic characteristics that are manifested in changes in speech prosody - these changes may, in turn, indicate an underlying mental condition. We have developed automatic computational tools for the monitoring of pathological mental states - including characterization, detection, and classification. We show that some features strongly correlate with perceptual diagnostic evaluation scales of both schizophrenia and depression, suggesting the contribution of such acoustic speech properties to the perception of an apparent mental condition. We further show that one can use these temporal domain features to correctly classify up to 87.5 % and up to 70 % of the speakers in a two-way and in a three-way classification tasks respectively.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
While it is possible to use existing automatic systems to produce a high dimensional non-specific description of the voice signal, we focus on a small set of meaningful features for two reasons: (i) These features appear to be ecologically relevant and correspond with psychiatrists’ intuition about the characteristic features of the speech of Schizophrenia patients. (ii) Our application domain suffers from the problem of small sample, which necessitates the use of low dimensional representations to enable effective learning; this is accomplished by choosing a small set of relevant features. The alternative, which is to use a high dimensional representation followed by dimensionality reduction (like PCA), typically leads to the unfortunate outcome that the final result is hard to interpret in terms of the underlying features.
- 2.
Note that with only three treatment groups, it’s overly conservative to adjust the alpha levels with a Bonferroni method as with only 3 treatment groups, there is little risk in an increasing Type I error rate.
References
Michaelis, D., Fröhlich, M., Strube, H.: Selection and combination of acoustic features for the description of pathologic voices. J. Acoust. Soc. Am. 103, 1628–1639 (1998)
Cohen, A., Alpert, M., Nienow, T., Dinzeo, T., Docherty, N.: Computerized measurement of negative symptoms in schizophrenia. J. Psychiatr. Res. 42, 827–836 (2008)
Moore, E., Clements, M., Peifer, J., Weisser, L.: Critical analysis of the impact of glottal features in the classification of clinical depression in speech. IEEE Trans. Biomed. Eng. 55, 96–107 (2008)
Alpert, M., Pouget, E., Silva, R.: Reflections of depression in acoustic measures of the patient’s speech. J. Affect. Disord. 66, 59–69 (2001)
Alpert, M., Shaw, R., Pouget, E., Lim, K.: A comparison of clinical ratings with vocal acoustic measures of flat affect and alogia. J. Psychiatr. Res. 36, 347–353 (2002)
American Psychiatric Association, American Psychiatric Association Task Force on DSM-IV: Diagnostic and Statistical Manual of Mental Disorders: DSM-IV-TR. American Psychiatric Publishing, Inc., Arlington (2000)
Andreasen, N.: Scale for the assessment of negative symptoms (sans). Br. J. Psychiatr. (1989)
Andreasen, N.: Negative symptoms in schizophrenia: definition and reliability. Arch. Gen. Psychiatr. 39, 784–788 (1982)
Andreasen, N., Alpert, M., Martz, M.: Acoustic analysis: an objective measure of affective flattening. Arch. Gen. Psychiatr. 38, 281–285 (1981)
Pennebaker, J., Mehl, M., Niederhoffer, K.: Psychological aspects of natural language use: our words, our selves. Ann. Rev. Psychol. 54, 547–577 (2003)
Newman, S., Mather, V.: Analysis of spoken language of patients with affective disorders. Am. J. Psychiatr. 94, 913–942 (1938)
Ishak, W.: Outcome measurement in psychiatry: a critical review. American Psychiatric Pub., Arlington (2002)
Kring, A., Bachorowski, J.: Emotions and psychopathology. Cogn. Emot. 13, 575–599 (1999)
Dsm, I.: Diagnostic and Statistical Manual of Mental Disorders. American psychiatric association, Arlington (1994)
Kay, S., Flszbein, A., Opfer, L.: The positive and negative syndrome scale (panss) for schizophrenia. Schizophr. Bull. 13, 261 (1987)
Montgomery, S., Asberg, M.: A new depression scale designed to be sensitive to change. Br. J. Psychiatr. 134, 382–389 (1979)
Hamilton, M.: A rating scale for depression. J. Neurol. Neurosurg. Psychiatr. 23, 56 (1960)
MATLAB: version 7.11.0 (R2010b). The MathWorks Inc., Natick, MA (2010)
Fan, R., Chang, K., Hsieh, C., Wang, X., Lin, C.: Liblinear: a library for large linear classification. J. Mach. Learn. Res. 9, 1871–1874 (2008)
Alpert, M., Kotsaftis, A., Pouget, E.: Speech fluency and schizophrenic negative signs. Schizophr. Bull. 23, 171–177 (1997)
Alpert, M., Rosenberg, S., Pouget, E., Shaw, R.: Prosody and lexical accuracy in flat affect schizophrenia. Psychiatr. Res. 97, 107–118 (2000)
Fan, R., Chang, K., Hsieh, C., Wang, X., Lin, C.: LIBLINEAR: a library for large linear classification. J. Mach. Learn. Res. 9, 1871–1874 (2008)
Acknowledgements
This work was supported in part by the Intel Collaborative Research Institute for Computational Intelligence (ICRI-CI), and the Gatsby Charitable Foundations.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing Switzerland
About this paper
Cite this paper
Kliper, R., Portuguese, S., Weinshall, D. (2016). Prosodic Analysis of Speech and the Underlying Mental State. In: Serino, S., Matic, A., Giakoumis, D., Lopez, G., Cipresso, P. (eds) Pervasive Computing Paradigms for Mental Health. MindCare 2015. Communications in Computer and Information Science, vol 604. Springer, Cham. https://doi.org/10.1007/978-3-319-32270-4_6
Download citation
DOI: https://doi.org/10.1007/978-3-319-32270-4_6
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-32269-8
Online ISBN: 978-3-319-32270-4
eBook Packages: Computer ScienceComputer Science (R0)