Friday, Oct. 28
10:45 – 11:45 (1 hr)
Meeting Room: Sheraton Hall C (Lower Concourse)
Bill Simpson*, PhD; Marlon Danilewitz, MD; Sunny Tang, MD; Anthony Yeung, MD

CanMEDS Roles:

1. Medical Expert
2. Scholar
3. Professional

At the end of this session, participants will be able to: 1) Review the fundamentals of artificial intelligence and natural language processing as it applies to clinical care; 2) Outline the current research to date involving natural language processing and schizophrenia; and 3) Discuss the potential implications, current barriers, and future directions of research in this domain.

Clinical evaluation of speech is a fundamental aspect of all psychiatric assessments, typically incorporated as part of the mental status exam—the psychiatrist’s physical exam. Although highly informative, clinical judgements of speech are subjective and lack precision. At the same time, advances in artificial intelligence technologies, such as quantitative voice analysis and natural language processing (NLP), have drastically improved our ability to objectively detect and characterize alterations in speech and language and have growing applications for people with psychiatric disorders. In particular, NLP has shown tremendous clinical potential in schizophrenia spectrum disorders (SSDs), where it has demonstrated strong evidence in predicting the onset of psychosis and distinguishing people with psychosis from control subjects.

Over the course of this interactive workshop, we will provide participants with an overview of the role of NLP as an emerging tool to understand speech and language biomarkers of SSDs. Participants will acquire an understanding of the fundamentals of NLP as a general process, the evidence supporting the use of NLP to diagnose SSD and track clinical symptoms, and implications for integration with current clinical care models. The panel will also highlight limitations of NLP in the current literature and outline emerging research to address these limitations. As a part of the workshop, we will integrate case-based learning and allow for ongoing audience engagement throughout the presentation.

References:

1. Corcoran CM, Mittal VA, Bearden CE, et al. Language as a biomarker for psychosis: a natural language processing approach. Schizophr Res 2020;226:158–66.
2. Parola A, Simonsen A, Bliksted V, et al. Voice patterns in schizophrenia: a systematic review and Bayesian meta-analysis. Schizophr Res 2020;216:24–40.