Deconstructing Psychoses Based on Patterns of Abnormal Brain Activity

描述（由申请人提供）：精神分裂症（SZ）和双相情感障碍（BD）这两种主要的精神障碍，在世界范围内继续被列为致残的主要原因，主要是因为目前的临床症状定义不足以治疗和预后，因为它们与潜在的病理生理学不充分一致。本提案使用研究领域标准（RDoC）框架来定义和验证精神病的生物学信息和临床相关的神经表型。具体来说，SZ和BD患者的神经影像学研究表明，与感知、认知控制和面部情感处理的RDoC领域相关的神经网络的连通性障碍是精神病病理生理学的核心。此外，这些结构域的异常被认为可以解释与精神障碍相关的临床症状和认知缺陷。因此，我们的总体假设是，在感知、认知控制和面部情感识别领域的通用神经网络中，有效连接的异常将检测到精神病的生物学和临床相关神经表型。我们提供的初步数据表明，SZ或BD患者可以根据其神经网络结构划分为亚组，并且这些神经表型可以映射到临床维度。我们的研究结果是基于在工作记忆过程中参与感知和认知控制的域通用网络的动态因果模型对有效连通性的估计。我们发现的神经表型在SZ和BD之间部分重叠，并与症状严重程度和临床病程相关。基于这一证据，本研究的目的是(a)扩展我们的初步结果，以便基于感知、认知控制和面部情感处理领域通用网络的动态因果模型得出的有效连接参数来确定精神病的神经表型，并在两个独立样本中测试其可重复性；(b)定义已确定的神经表型与症状学和病程的临床维度之间的关联。(c)确定其对治疗反应的预测价值。该建议得益于动态因果模型的使用，该模型可以推断出潜在网络动态变化的大脑区域之间的因果相互作用，通过基于结果的可重复性测试我们的结果的有效性，并评估已确定的神经表型的治疗相关性。本申请所提出的研究的成功完成将提高我们对精神病中脑连接异常的临床和预后意义的理解，为治疗计划提供科学依据，促进有针对性的病因调查和新的治疗方法的开发。

DESCRIPTION (provided by applicant): The main psychotic disorders, schizophrenia (SZ) and bipolar disorder (BD), continue to rank amongst the leading causes of disability worldwide largely because current clinical syndromal definitions are insufficient for treatment and prognosis because they are inadequately aligned with underlying pathophysiology. This proposal uses the Research Domains Criteria (RDoC) framework in order to define and validate biologically informed and clinically relevant neural phenotypes for psychotic disorders. Specifically, neuroimaging studies in patients with SZ and BD suggest that dysconnectivity within neural networks linked to the RDoC domains of perception, cognitive control and facial affect processing is central to the pathophysiology of psychosis. Further, abnormalities in these domains have been proposed to explain the clinical symptoms and cognitive deficits associated with psychotic disorders. Accordingly, our overall hypothesis is that abnormalities in effective connectivity, within domain-general neural networks of perception, cognitive control and facial affect identification, will detect biologically and clinically relevant neural phenotypes for psychosis. We present preliminary data that show that patients with SZ or BD can be classified into subgroups defined by their neural network architecture and that these neural phenotypes can be mapped onto clinical dimensions. Our results are based on estimates of effective connectivity from a dynamic causal model of the domain-general networks engaged in perception and cognitive control during working memory. The neural phenotypes we identified showed partial overlap between SZ and BD and were associated with symptom severity and clinical course. Based on this evidence, the aims of this proposal are (a) to expand our preliminary results in order to identify neural phenotypes for psychosis based on effective connectivity parameters derived from dynamic causal models of domain-general networks of perception, cognitive control and facial affect processing and test their reproducibility in two independent samples, (b) to define the association between the identified neural phenotypes and clinical dimensions of symptomatology and course, and (c) to determine their predictive value for treatment response. The proposal benefits from the use of dynamic causal modelling, which can infer causal interactions between brain regions underlying altered network dynamics, from testing the validity of our results based on their reproducibility and from assessing the therapeutic relevance of the identified neural phenotypes. Successful completion of the studies proposed in this application will improve our understanding of the clinical and prognostic significance of abnormal brain connectivity in psychosis, provide a scientific basis for therapeutic planning, and facilitate targeted etiological investigations and the development of new therapeutic approaches.