Neural Network Predictors of Treatment Outcome Among Adolescent Assault Victims

描述(由申请人提供)：NIMH探索/发展助学金奖(R21)的建议旨在确定女性青少年攻击受害者中与创伤为重点的认知行为治疗(TF-CBT)反应相关的神经功能连接模式。青少年遭受攻击是持续精神病理的一个潜在风险因素，最明显的是创伤后应激障碍。对于青少年创伤后应激障碍患者，TF-CBT是唯一具有强大经验支持的治疗方法，然而对TF-CBT的反应是不同的，许多患者在治疗后继续表现出明显的临床症状。这项建议的总体目标是使用计算神经科学工具来预测和了解这一弱势群体的治疗反应。基于人类神经成像研究表明，神经网络中调节创伤后应激障碍受害者的情绪反应性和情绪调节的活性和连通性改变，我们假设这些神经网络中的功能连通性模式可以用于预测和理解青少年攻击受害者对TF-CBT的反应。45名年龄在11-16岁的青少年暴力受害者将接受为期12周的TF-CBT课程。参与者将在治疗前后进行情绪反应和情绪调节任务时进行fMRI扫描。结合图论分析和支持向量分类与回归，将被用来识别预测TF-CBT后续反应的功能连通性的前处理模式(目标1)。图论分析同样将用于确定网络组织从治疗前到治疗后的变化，这些变化与治疗成功(目标2)和不成功(目标3)有关。这种对了解TF-CBT可变反应的临床问题的分析方法将促进临床医生用来预测儿童治疗反应的具体算法，这是为这一脆弱群体实现个性化治疗的第一步。此外，这种分析方法将确定调节治疗反应的基本神经机制，并为新治疗成分的开发提供靶点。这一应用为了解弱势青少年群体的治疗反应提出了一种新的方法，并有望促进制定更一致的干预措施，以减轻与青少年攻击暴露相关的高昂成本。

DESCRIPTION (provided by applicant): This proposal for a NIMH Exploratory/Developmental Grant Award (R21) seeks to identify neural functional connectivity patterns associated with response to Trauma-Focused Cognitive Behavioral Therapy (TF-CBT) among female adolescent assault victims. Adolescent assault exposure is a potent risk factor for persistent psychopathology, most notably PTSD. TF-CBT is the only treatment for adolescent PTSD victims with strong empirical support, yet response to TF-CBT is variable and many victims continue to exhibit clinically significant symptoms following treatment. The overall goal of this proposal is to use computational neuroscience tools to predict and understand treatment response among this vulnerable population. Based on human neuroimaging studies demonstrating altered activity and connectivity within neural networks mediating emotion reactivity and emotion regulation among PTSD victims, we hypothesize that patterns of functional connectivity within these neural networks can be used to predict and understand response to TF-CBT among adolescent assault victims. 45 adolescent assault victims aged 11-16 will be provided with a 12-week course of TF-CBT. Participants will undergo fMRI scanning while engaged in emotion reactivity and emotion regulation tasks before and after treatment. A combination of graph theory analyses and support vector classification and regression will be used to identify pre-treatment patterns of functional connectivity that predict subsequent response to TF-CBT (Aim 1). Graph theory analyses will similarly be used to identify changes in network organization from pre-to-post-treatment associated with successful (Aim 2) and unsuccessful (Aim 3) treatment response. This analytic approach to the clinical problem of understanding the variable response to TF-CBT will foster concrete algorithms to be used by a clinician to predict a child's treatment response, which is the first step towards personalizing treatments for this vulnerable population. Further, this analytic approach will identify the essential neural mechanism mediating treatment response and provide targets for the development of novel treatment components. This application proposes a novel approach towards understanding treatment response among a vulnerable adolescent population and will hopefully facilitate the development of more consistent interventions to ameliorate the high cost associated with adolescent assault exposure.