Description: You drop into your doctor’s office and tell them you are not feeling well. Assuming COVID is not the issue (which would likely be picked up at the front desk of the clinic using screening questions) what is one of the first things the doctor will likely do? Take you temperature, right? (Or ask you if you have a fever if you connect with them on the phone or via Zoom). Why do physicians do this? Well because your temperature, if it is elevated above “normal,” is an indicator of something being physically amiss with you. Now, what if you contact your therapist and tell them you are not “feeling right”? What would be the fist thing they do? Well, whatever it, is would very likely NOT involve a thermometer. It would likely involve a lot of questions about how you are feeling, what has been happening with you lately how your relationships are going, what has changed for you recently, and perhaps how long have you been experiencing these feelings? Wouldn’t it be helpful if there was a general, quick, test that could be done that would identify the presents of mental unwell-ness and perhaps even suggest what the issue might be? What might such a test involve? Well, for a possibility, have a look at this video. Really, go and look at the video and then come back here. Now, besides the issue with the character in the video having always been like that (sounding like that), what are some other issues that might be at play if one was to try and develop an AI app that would be aimed at identifying mental state and mental health through voice analysis? Think about that for a minute and then read the article linked below for a discussion of those sorts of questions and references to a few sample apps.
Source: Can A.I – Driven Voice Analysis Help Identify Mental Disorders? Ingrid K. Williams, The New York Times.
Date: April 5, 2022
So, do you think there is promise in voice samples being the body temperature indicator of mental health or change in mental health? Clearly more research — research — is needed. As well, we will need to get quite a distance beyond simply ascribing depressive symptoms to anyone who sounds like Eeyore. However, there ARE somewhat typical physiological issues correlated with particular disorders. The unmodulated, slower, Eeyore speech of people who are depressed or the tense somewhat sped up speech of people who are anxiety ridden. So, perhaps if voice sample assessment could be done regularly and compared to track individual changes (something that did not seem to be part of the Apps discussed in the linked article) maybe this could lead to a mental health temperature indicator.
Questions for Discussion:
- What are some examples (maybe just stereotypes) of disorder/voice combinations?
- What sorts of research needs to be done to see if voice sampling could work as a screen for mental health issues?
- If research into this starts to look promising, how should the use of such apps be set up in order to both continue gathering validity data and to tune them for use at the individual level?
References (Read Further):
Gravenhorst, F., Muaremi, A., Bardram, J., Grünerbl, A., Mayora, O., Wurzer, G., … & Tröster, G. (2015). Mobile phones as medical devices in mental disorder treatment: an overview. Personal and Ubiquitous Computing, 19(2), 335-353. Link
Parola, A., Simonsen, A., Bliksted, V., & Fusaroli, R. (2020). Voice patterns in schizophrenia: A systematic review and Bayesian meta-analysis. Schizophrenia research, 216, 24-40. Link
Scherer, S., Stratou, G., Gratch, J., & Morency, L. P. (2013, August). Investigating voice quality as a speaker-independent indicator of depression and PTSD. In Interspeech (pp. 847-851). Link
Mundt, J. C., Snyder, P. J., Cannizzaro, M. S., Chappie, K., & Geralts, D. S. (2007). Voice acoustic measures of depression severity and treatment response collected via interactive voice response (IVR) technology. Journal of neurolinguistics, 20(1), 50-64. Link
Cannizzaro, M., Harel, B., Reilly, N., Chappell, P., & Snyder, P. J. (2004). Voice acoustical measurement of the severity of major depression. Brain and cognition, 56(1), 30-35. Link
Hashim, N. W., Wilkes, M., Salomon, R., Meggs, J., & France, D. J. (2017). Evaluation of voice acoustics as predictors of clinical depression scores. Journal of Voice, 31(2), 256-e1. Link
Cohen, A. S., Fedechko, T. L., Schwartz, E. K., Le, T. P., Foltz, P. W., Bernstein, J., … & Elvevåg, B. (2019). Ambulatory vocal acoustics, temporal dynamics, and serious mental illness. Journal of abnormal psychology, 128(2), 97. Link