'

Using technology and data-driven systems to help detect signs of mental distress with Dr. Rebecca Resnik and Dr. Philip Resnik

Wednesday, October 21, 2020

Philip Resnik, PhD, returns to the Psychcast, this time with his research partner and wife, Rebecca Resnik, PsyD, to discuss the interface between language, psychiatry, psychology, and health.

Dr. Philip Resnik appeared on the show previously to discuss artificial intelligence, natural language processing, and mental illness. He is a professor in the department of linguistics at the University of Maryland, College Park, and has a joint appointment with the university’s Institute for Advanced Computer Studies.

Dr. Philip Resnik has disclosed being an adviser for Converseon, a social media analysis firm; FiscalNote, a government relationship management platform; and SoloSegment, which specializes in enterprise website optimization. Some of the work Dr. Philip Resnik discusses has been supported by an Amazon AWS Machine Learning Research Award.

Dr. Rebecca Resnik is a licensed psychologist in private practice who specializes in neuropsychological assessment. In 2014, she served as cofounder of the Computational Linguistics and Clinical Psychology workshop at the North American Association for Computational Linguistics. She continues to serve as a workshop organizer and clinical consultant to the cross-disciplinary community. She has no disclosures.

Dr. Norris disclosed having no conflicts of interest.

Take-home points

  • Dr. Rebecca Resnik and Dr. Philip Resnik are interested in finding measurable, observable features to apply to the assessment of psychological and psychiatric diagnoses. They point out that finding an objective measure is essential for scaling up mental health evaluations and treatment.
  • Natural language processing (NLP) is focused on analyzing language content. NLP technology has generated tools such as Siri, Alexa, and Google Translate, and NLP allows computers to do things more intelligently with human language.
  • Individuals are using machine learning and NLP to analyze language data sets to evaluate diagnostic criteria. The goal is to create or use language sets that can be analyzed outside of the clinic.
  • Dr. Rebecca Resnik imagines a world where a patient gives a “language sample” to an app or an avatar that would be evaluated by NLP that would, in turn, offer some overarching hypotheses about the person. So much of evaluations is trying to home in on the correct signal, explicit and implicit, from the patient. In addition, neuropsychiatric tests/scales are standardized against a limited scope of the population, so NLP would be matched to the individual.
  • Dr. Philip Resnik looks at signals in text and speech content, acoustics, microexpressions, and even biometric data. Machine learning can process and distill a huge amount of data with various signals more easily than any human.
  • Dr. Rebecca Resnik revisits the idea of clinical white space, which is the “space” or the time between clinical encounters, and this is where decompensation and high-risk suicidal behaviors occur. She suggests that NLP software could be used to fill this white space by using apps to collect text samples from patients, and the software would analyze the samples and warn of patients who are at risk of decompensation or suicide. If clinicians were to use text or speech samples from people’s smart technology, we could assess an individual's risk in the moment and use nudge-type interventions to prevent suicide.
  • Finally, Dr. Philip Resnik emphasizes that there are technologists who have the skills and technology that is on the verge of helping clinicians, but the key to progress is collaborating with clinicians.

References

Resnik P et al. J Analytical Psychol. 2020 Sep 10. doi: 10.111/sltb.12674.

Coppersmith G et al. Biomed Inform Insights. 2018;10:1178222618792860.

Zirikly A et al. CLPsych 2019 shared task: Predicting the degree of suicide risk in Reddit posts. Proceedings of the Sixth Workshop on Computational Linguistics and Clinical Psychology. 2019 Jun 16.

Yoo DW et al. JMIR Mental Health. 2020;7(8):e16969.

American Medical Informatics Association and Mental Health: https://www.amia.org/mental-health-informatics-working-group

Selanikio J. The big-data revolution in health care. TEDxAustin. 2013 Feb.

CLPsych: Computational Linguistics and Clinical Psychology Workshop. 2019 Program.

* * *

Show notes by Jacqueline Posada, MD, associate producer of the Psychcast; assistant clinical professor in the department of psychiatry and behavioral sciences at George Washington University in Washington; and staff physician at George Washington Medical Faculty Associates, also in Washington. Dr. Posada has no conflicts of interest.

* * *

For more MDedge Podcasts, go to mdedge.com/podcasts

Email the show: [email protected]

Podcast Participants

Lorenzo Norris, MD
Lorenzo Norris, MD, is host of the MDedge Psychcast, editor in chief of MDedge Psychiatry, and assistant professor of psychiatry and behavioral sciences at George Washington University, Washington. He also serves as assistant dean of student affairs at the university, and medical director of psychiatric and behavioral sciences at GWU Hospital. Dr. Lorenzo Norris has no conflicts of interest.
Renee Kohanski, MD
Renée S. Kohanski, MD, is a board-certified psychiatrist with additional training in forensic psychiatry. She has been a board examiner for the American Board of Psychiatry and Neurology, and has enjoyed a broad-based practice in academic, community, and forensic psychiatry. She is currently a solo practitioner and owner of RK Psychiatry Associates and serves on the Editorial Advisory Board of MDEdge Psychiatry. Talkers magazine describes Dr. Kohanski as “one of the most reliable ‘go-to’ sources for insights and information about psychiatry in the media today.” Dr. Renée Kohanski has no conflicts of interest.