Collaborative Research: Development of a precision closed loop BCI for socially fearful teens with depression and anxiety

该项目旨在开发创新技术，帮助治疗师在与患有焦虑或抑郁的青少年进行谈话治疗期间评估和康复社交恐惧。根据世界卫生组织的数据，抑郁症和焦虑症是全球残疾的主要原因，影响了大约25%的人口，每年给全球经济造成1万亿美元的损失。治疗青少年焦虑或抑郁的标准方法是每周与治疗师进行多次一对一的暴露治疗。暴露疗法逐渐将青少年暴露在引发他们恐惧的现实社会环境中，同时为他们提供管理和容忍他们痛苦的工具。高达50%的青少年对这种治疗方法没有反应，使他们面临慢性症状、自杀、残疾和显著缩短预期寿命的高风险。治疗失败的部分原因是很难在治疗过程中重现现实生活中的社会挑战，因此青少年无法在治疗师的直接监督下面对恐惧进行练习。目前还没有专门设计用于在暴露治疗期间评估、重建和恢复社会恐惧的商业产品。该项目由一个多学科团队领导，旨在通过开发增强现实(AR)技术来应对这些挑战，该技术可以创建在治疗环境中唤起社会恐惧的模拟现实情景，以及社会恐惧的标记，这些标记将用于在青少年及其治疗师在治疗期间练习技术时向他们提供客观和可操作的反馈。为了实现这些目标，该项目将包括与社区专家合作，并为K-12到研究生水平提供多学科培训机会，重点是纳入边缘社区的受训人员。该项目的研究目标是为社交恐惧的青少年介绍一种AR引导的基于脑电(EEG)的暴露技术的临床应用原型。建议的技术将：(I)使用新的硬件集成和软件开发来无缝地同步脑电数据采集和现实生活中人际关系中社交恐惧情景的AR呈现；(Ii)通过脑电特征选择和贝叶斯推理方法(同时考虑个体反应和总体反应)准确和连续地检测青少年是否表现出恐惧反应；(Iii)设计新的机器学习方法来识别基于EEG的个性化恐惧指数；(Iv)实时监测基于EEG的恐惧指数，并在必要时调整AR社会恐惧情景以增加恐惧程度；以及(V)识别个人化阈值以检测用户何时处于“曝光区”中，并提供视觉反馈以提示青少年何时应用治疗师所规定的特定曝光技术。拟议的系统有可能为暴露疗法提供一个技术驱动的范例，有意义地反映抑郁和焦虑的青少年所经历的社会挑战。贝叶斯最优统计推理将支持这项技术，提供数学驱动的框架，提高EEG记录的准确性，并结合基于AR的头部跟踪。结果将在同行评议的文章、外展计划和代码/数据存储库中公开发布。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This project aims to develop innovative technology that can assist therapists in assessing and rehabilitating social fears during talk therapy with teens suffering from anxiety or depression. According to the World Health Organization, depression and anxiety are leading causes of disability worldwide, affecting approximately 25% of the population and costing the global economy $1 trillion per year. The standard approach to treating teen anxiety or depression involves many weekly sessions of one-on-one exposure therapy with a therapist. Exposure therapy gradually exposes teens to real-life social situations that trigger their fears while providing them with tools to manage and tolerate their distress. Up to 50% of teens do not respond to this treatment approach, placing them at high risk for chronic symptoms, suicidality, disability, and a significantly shorter life expectancy. Treatment failure occurs partially because it is hard to recreate real-life social challenges within therapy sessions, so teens are not able to practice facing their fears under the direct supervision of their therapists. There is currently no commercially available product specifically designed to assess, recreate, and rehabilitate social fears during exposure therapy. This project, led by a multidisciplinary team, aims to address these challenges by developing augmented reality (AR) technology that creates simulated real-life situations that evoke social fears within the therapy environment, and markers of social fear that will be used to provide objective and actionable feedback to teens and their therapists as they practice techniques during therapy sessions. To achieve these goals, the project will include partnerships with community experts and offer multidisciplinary training opportunities for K-12 to graduate level, with an emphasis on inclusion of trainees from marginalized communities.The research objective of this project is to introduce a prototype for clinical application of an AR-guided, electroencephalogram (EEG)-based exposure technology for socially fearful teens. The proposed technology will: (i) use novel hardware integration and software development to seamlessly synchronize EEG data acquisition with AR presentation of social fear scenarios during real-life interpersonal situations; (ii) accurately and continuously detect whether the teen is exhibiting fearful vs. not fearful responses through EEG feature selection and Bayesian inference methods (considering both individualized responses and generalized responses across population); (iii) design novel machine learning methods to identify individualized EEG-based fear indices; (iv) monitor EEG-based fear indices in real time and adjust the AR social fear scenarios to increase the level of fear when necessary; and (v) identify individualized thresholds to detect when the user is in the “exposure zone” and provide visual feedback to prompt the teen when to apply specific exposure techniques as prescribed by the therapist. The proposed system has the potential to provide a technology-driven paradigm for exposure therapy that meaningfully reflects the social challenges experienced by depressed and anxious teens. Bayesian optimal statistical inference will support the technology, providing mathematically-driven frameworks that enhance the accuracy of EEG recordings coupled with AR-based headtracking. Outcomes will be openly disseminated in peer-reviewed articles, outreach programs, and code/data repositories.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.