Symposium VI: Advances on the Digital Frontier Speakers
Dror Ben-Zeev
Ditte Lammers Vernal
Justin Tauscher
Ben Buck
Sarah Kopelovich
Title: Development and Evaluation of FOCUS: A Smartphone Intervention for Serious Mental Illness
Abstract: mHealth approaches that use smartphones to deliver interventions can play an important role in improving access to care for people with serious mental illness. This presentation will describe the systematic development and comparative effectiveness trial of FOCUS: a recovery-oriented CBTp informed smartphone intervention. FOCUS treatment targets include auditory hallucinations, mood difficulties, sleep dysregulation, medication use, and social functioning. Following a staged iterative design and development process involving people with lived experience and providers, we conducted a randomized controlled trial of FOCUS vs. a clinic-based group intervention. In the talk I will review the clinical findings of the trial and provide an update on a large scale implementation study of FOCUS currently underway in Washington State.
Title: Challenge - Virtual Reality for Auditory Hallucinations
Abstract: In the Danish RCT-study Challenge, we investigated virtual reality-assisted therapy (VRT) for auditory hallucinations in adults with psychosis. Pending finalization by June 2024, the results will be presented. Otherwise, the presentation will summarize the study's design and feature a compelling case: a young man with comorbid schizophrenia, autism, and learning disabilities. He initially benefited from VRT but later relapsed before reinitiating the therapy. Qualitative interviews with the patient, contact-person, and therapist enrich our findings. Currently, we are planning a youth feasibility study followed by an RCT, and the presentation will briefly outline proposed therapy and VR adjustments for youth.
Title: Automated Detection of Thinking Styles and Cognitive Distortions in Psychosis and Serious Mental Illness
Abstract: This presentation provides an overview of research utilizing natural language processing (NLP) to detect and classify distorted thinking in people at risk for psychosis. We demonstrate effectiveness in clinical text-message exchanges from individuals with serious mental illness and audio diaries from people with auditory verbal hallucinations. We discuss limitations of current approaches, novel methodologies to enhance accuracy of the classification process, and evaluate the utility of a large language model in this context. We show that NLP can be a valuable tool in identifying cognitive distortions, offering a scalable and efficient approach to enhance clinical decision-making and mental health intervention.
Title: The Bolster Project: mHealth for Caregivers to Youth at Risk for Psychosis
Abstract: Family caregivers have a significant impact on treatment engagement and outcomes for youth at risk for psychosis. They are often the first to recognize symptoms and seek out treatment on behalf of the affected person. Despite the significant impact of the duration of untreated illness, youth at risk for psychosis and their families face multiple varied barriers to mental health services. Digital interventions have demonstrated promise in expanding the reach of services for families affected by psychosis, but few self-guided tools have been developed for caregivers, and even fewer have been designed for the early help-seeking period. Our team developed a self-guided mHealth intervention designed to provide psychoeducation, communication coaching and wellness support to caregivers to youth at risk for psychosis. In this presentation, we introduce our user-centered design process, present field trial results, and discuss next steps for an ongoing clinical trial examining this intervention.
Title: Investigation of the First Artificial Intelligence-Based Cognitive Behavioral Therapy Training Tool
Abstract: The accessibility of training and fidelity assessment is critical to implementing and sustaining Evidence-Based Psychotherapeutic Interventions like CBTp. CBTpro is an asynchronous CBTp training tool that provides Artificial Intelligence (AI)-enabled feedback to learners across eight discrete CBTp skills. An iterative end user-informed development and validation process resulted in a functional tool that provides multimodal asynchronous training followed by standardized behavioral rehearsal tasks of high-yield cognitive and behavioral techniques for psychosis. This presentation will demonstrate the training tool’s functionality, present machine learning validation data, and will present preliminary data from this NIMH-funded Research & Development grant. Results suggest strong potential for AI to support standardized, automated training and clinical quality assurance across large systems of care. Additional studies, currently underway, will address research questions related to CBTpro’s effectiveness in shaping practitioner behavior with real clients, client outcomes, and the facilitators and barriers to asynchronous AI-enabled training.