JMIR Publications

ABSTRACT

Background:

Computational psychiatry and digital phenotyping have the potential to revolutionize mental health research and treatment development. Studies using these methodologies demonstrate that it is possible to identify and predict mental health episodes in psychiatric samples. A complementary approach aims to improve mental health risk assessment in the general population. This approach requires collecting data that are guided by a transdiagnostic and dynamic understanding of mental health, are well-suited to cutting-edge machine learning methods, and can support interdisciplinary collaboration.

Objective:

We aim to create the first computational psychiatry data set that includes rich, sensor-based, behavioral features and is optimized for both mental health and machine learning considerations. This data set is designed to be shared across academia, industry, and government using gold standard guidelines for privacy, confidentiality, and data integrity.

Methods:

We are using a stratified, random sampling design with two crossed factors (difficulties with emotion regulation and perceived life stress) to recruit a sample of 400 adults balanced across high- and low-risk for mental health conditions. Participants first complete self-report questionnaires assessing current and lifetime psychiatric and medical diagnosis and treatment, and current psychological symptoms, social functioning, and substance use. Participants then complete a 7-day in situ phase of data collection that includes providing daily audio recordings, passive sensor data collected from Smartphones, self-reports of daily mood and important events, and a verbal description of the most positive and negative events of their day during a nightly phone call. Participants complete the same baseline questionnaires at 6- and 12-months after the 7-day data collection phase. Self-reported medical and psychiatric encounters will be verified and augmented by medical records. We plan to use machine learning techniques to determine the likelihood of experiencing a mental health event during the 6- and 12-month follow-up periods based on data collected during baseline and the 7-day study phase.

Results:

Data collection began in June 2022 and is expected to conclude July 2024. To date, 310 participants have consented to the study, 149 have completed the baseline questionnaire and 7-day intensive data collection phase, and 61 and 31 have completed the 6- and 12-month follow-up questionnaires, respectively. Once completed, the proposed data set will be made available to academic researchers, industry, and government.

Conclusions:

This data set is designed to advance the field’s ability to identify community-dwelling individuals at risk for a future mental health event. This is a complementary approach to current digital phenotyping methods and computational psychiatry research that has the potential to improve mental health risk assessment within the general population. This data set aims to support the field’s move away from siloed research labs collecting proprietary data, and towards interdisciplinary collaborations that incorporate clinical, technical, and quantitative expertise at all stages of the research process.