BROWSE
Detail
Journal Article
Development of prediction models for screening depression and anxiety using smartphone and wearable-based digital phenotyping: protocol for the Smartphone and Wearable Assessment for Real-Time Screening of Depression and Anxiety (SWARTS-DA) observational study in Korea
- Authors
- Yu-Bin Shin, Ah Young Kim, Seonmin Kim, Min-Sup Shin, Jinhwa Choi, Kyung Lyun Lee, Jisu Lee, Sangwon Byun, Sujin Kim, Heon-Jeong Lee, Chul-Hyun Cho
- Issue Date
- 2025-06
- Citation
- BMJ Open, v.15, pp.1-11
- ISSN
- 2044-6055
- Publisher
- BMJ Publishing Group
- Language
- English
- Type
- Journal Article
- Abstract
- Introduction Depression and anxiety are highly prevalent mental health conditions that significantly affect quality of life and cause societal burdens. However, their detection and diagnosis rates remain low owing to the limitations of the current screening methods. With rapid technological advancements and the proliferation of consumer-grade wearable devices and smartphones, their integration into digital phenotyping research has enabled the unobtrusive screening for depression and anxiety in natural settings. The Smartphone and Wearable Assessment for Real-Time Screening of Depression and Anxiety study aims to develop prediction algorithms to identify individuals at risk for depressive and anxiety disorders, as well as those with mild-to-severe levels of either condition or both. By collecting comprehensive data using smartphones and smartwatches, this study aims to facilitate the translation of artificial intelligence-based early detection research into clinical impact, thereby potentially enhancing patient care through more accurate and timely interventions. Methods and analysis This cross-sectional observational study will enrol up to 2500 participants (at least 1000) aged 19-59 years from South Korea via social media outreach and clinical referrals. The eligible participants must use a compatible smartphone. Each participant will be followed up for 4 weeks. Data will be collected using a custom-developed smartphone application called PixelMood. Active data collection will include daily, weekly and monthly self-report questionnaires incorporating validated scales, such as the Patient Health Questionnaire-9 and Generalized Anxiety Disorder-7. Passive data from smartphones include information on physical activity, location, ambient light and smartphone usage patterns. Optionally, participants using the Apple Watch or Galaxy Watch devices can provide additional data on physiological responses and sleep health. The primary outcome will be the development of machine-learning algorithms to predict depression and anxiety based on these digital biomarkers. We will employ various machine-learning techniques, including random forest, support vector machine and deep-learning models. The secondary outcomes will include the association between digital biomarkers and clinical measures, and the feasibility and acceptability of data collection methods. Various features characterising mobile usage behaviours, physical/social activity, sleep patterns, resting physiological states and circadian rhythms will be exploited to serve as potential digital phenotyping markers. Advanced machine-learning and deep-learning techniques will be applied to multimodal data for model generation. Ethics and dissemination This study protocol was reviewed and approved by the Institutional Review Board of the Korea University Anam Hospital (approval number: 2023AN0506). The results of this study will be disseminated via multiple channels. The findings will be presented at local, national and international conferences in relevant fields, such as psychiatry, psychology and digital health. Manuscripts detailing the study results will be submitted to peer-reviewed journals for publication.
- KSP Keywords
- Active data, Ambient light, Circadian rhythms, Clinical impact, Cross-sectional, Data Collection, Detection and diagnosis, Early detection, Generalized anxiety disorder(GAD), Institutional Review Board(IRB), Mobile usage
This work is distributed under the term of Creative Commons License (CCL)
(CC BY NC)
(CC BY NC)
