Feasibility Study for Estimation of Depression Severity using Voice Analysis

Voice analysis is suitable for screening large populations and daily monitoring of illnesses because of its advantages of cost and usability. We examined the feasibility of estimating the severity of major depressive disorder (MDD) using voice analysis. We developed an estimator for the Hamilton depression rating scale (HAM-D), which is the gold-standard of evaluation for MDD, by iteration of multiple regressions. As a result, the estimation index had a high correlation (0.832) with the HAM-D score. The results thus demonstrated the feasibility of voice-based screening and monitoring of MDD.

[1]  E. Paykel,et al.  Altered salivary dehydroepiandrosterone levels in major depression in adults , 2000, Biological Psychiatry.

[2]  Sandy Berger,et al.  Autonomy of Autonomic Dysfunction in Major Depression , 2009, Psychosomatic medicine.

[3]  W. Weintraub,et al.  THE APPLICATION OF VERBAL BEHAVIOR ANALYSIS TO THE STUDY OF PSYCHOLOGICAL DEFENSE MECHANISMS. IV: SPEECH PATTERN ASSOCIATED WITH DEPRESSIVE BEHAVIOR , 1964, The Journal of nervous and mental disease.

[4]  N. Herrmann,et al.  A Meta-Analysis of Cytokines in Major Depression , 2010, Biological Psychiatry.

[5]  M. Hamilton.  A RATING SCALE FOR DEPRESSION , 1960, Journal of neurology, neurosurgery, and psychiatry.

[6]  A. Beck,et al.  An inventory for measuring depression. , 1961, Archives of general psychiatry.

[7]  Björn W. Schuller,et al.  Recent developments in openSMILE, the munich open-source multimedia feature extractor , 2013, ACM Multimedia.

[8]  J A STARKWEATHER,et al.  VOICE QUALITY IN DEPRESSION. , 1965, Journal of abnormal psychology.

[9]  C. Hoge,et al.  Combat duty in Iraq and Afghanistan, mental health problems and barriers to care. , 2004, U.S. Army Medical Department journal.