Scalable and Quantifiable Mental Health Signals

Share this content:
Automated analysis can play a valuable role in improving patient mental health-related outcomes.
Automated analysis can play a valuable role in improving patient mental health-related outcomes.
 

There is always more that could be done for validation, but some of the possibilities provided by these advances bear enumerating. Using computational models of data gathered from patients outside of the clinical setting allows for real-time monitoring of their mental health and alerting when their communications reach a troubling state. It allows for population-level analysis at different geographic and temporal resolution than was previously possible (eg, “What was the reaction to Robin Williams' suicide?”).10,11

Most importantly, however, it enables a metric of mental health that can be measured without human intervention (by patient or clinician). When properly integrated into a patient's life and a clinician's workflow, it enables and empowers a wide range of data-driven approaches to psychiatry. When such metrics are paired with outcomes data on the efficacy of treatments, we may be able to predict the best intervention for a given patient at a given point in his or her life.

When paired with information about successful and unsuccessful patient-clinician relationships, we may be able to pair patients with a clinician based on the likelihood of successful outcomes. When these metrics are collected over the long term, we can characterize the long-term efficacy of care and support continued research into the effects of treatments over a patient's lifetime.

Empowering clinicians and peer-support personnel by scaling clinical impact has profound potential, and this is just the beginning.

Glen Coppersmith, PhD, is the founder and CEO of Qntfy, a company working to scale clinical impact and empower mental health professionals via technology, which designed and tested some of the algorithms mentioned in this article (for more, see http://qntfy.io). He holds degrees from Northeastern University in computer science and psychology, and was a research scientist at Johns Hopkins University from 2008 to 2014.

References

  1. Wang R, Chen F, Chen Z, et al. Studentlife: assessing mental health, academic performance and behavioral trends of college students using smartphones. In: Proceedings of the 2014 ACM International Joint Conference on Pervasive and Ubiquitous Computing.  New York, NY: Association for Computing Machinery; 2014;3-14.
  2. Resnik P, Resnik R, Mitchell M, eds. Proceedings of the Workshop on Computational Linguistics and Clinical Psychology: From Linguistic Signal to Clinical Reality. Baltimore, MD: Association for Computational Linguistics; 2014.
  3. Mitchell M, Coppersmith G, Hollingshead K, eds. Proceedings of the Workshop on Computational Linguistics and Clinical Psychology: From Linguistic Signal to Clinical Reality. Denver, CO: North American Association for Computational Linguistics; 2015.
  4. Chung C, Pennebaker J. The psychological functions of function words. In: Fiedler K, ed. Social Communication. New York, NY: Psychology Press; 2007:343-359.
  5. De Choudhury M, Gamon M, Counts S, Horvitz E. Predicting depression via social media. In: Proceedings of the 7th International AAAI Conference on Weblogs and Social Media (ICWSM). Palo Alto, CA: Association for the Advancement of Artificial Intelligence; 2013:128-137.
  6. Coppersmith G, Dredze M, Harman C, Hollingshead K. From ADHD to SAD: Analyzing the language of mental health on Twitter through self-reported diagnoses. In: Proceedings of the Workshop on Computational Linguistics and Clinical Psychology: From Linguistic Signal to Clinical Reality. Denver, CO: North American Chapter of the Association for Computational Linguistics: 2015.
  7. Coppersmith G, Dredze M, Harman C. Quantifying mental health signals in Twitter. In: Proceedings of the Workshop on Computational Linguistics and Clinical Psychology: From Linguistic Signal to Clinical Reality. Baltimore, MD: Association for Computational Linguistics; 2014.
  8. Coppersmith G, Harman C, Dredze M. Measuring posttraumatic stress disorder in Twitter. In: Proceedings of the 8th International AAAI Conference on Weblogs and Social Media (ICWSM). Palo Alto, CA: Association for the Advancement of Artificial Intelligence; 2014.
  9. Mitchell M, Hollingshead K, Coppersmith G. Quantifying the language of schizophrenia in social media. In: Proceedings of the Workshop on Computational Linguistics and Clinical Psychology: From Linguistic Signal to Clinical Reality. Denver, CO: North American Chapter of the Association for Computational Linguistics; 2015.
  10. Kumar M, Dredze M, Coppersmith G, De Choudhury M. Detecting changes in suicide content manifested in social media following celebrity suicides. In: Proceedings of the 26th ACM Conference on Hypertext and Hypermedia. New York, NY: Association for Computing Machinery; 2015.
  11. Coppersmith G. Quantified temporal and geographic mental health signals from social media. In: Proceedings of the Joint Statistical Meetings. Alexandria, VA: American Statistical Association; 2015.
Page 3 of 3
You must be a registered member of Psychiatry Advisor to post a comment.

Sign Up for Free e-newsletters