Advancing understanding of neural representations of threat perception through a novel predictive coding framework

项目总结 据估计，全球每年有6.7%至7.3%的人受到焦虑症的影响，并给 人们的生命。为了解决这个紧迫的问题，有必要更好地了解神经 焦虑的主观体验基础，包括威胁感知。翻译神经科学已经将注意力集中在 关于涉及核心区域的防御行为的动物模型。尽管这些动物模型和 他们启发的人类研究对象在治疗焦虑方面取得了进展，焦虑是神经之间的映射 机制和主观经验仍然知之甚少。找到支持的区域集 动物模型中的防御行为似乎并不涉及所有恐惧或焦虑的情况。这个 目前的项目通过将焦虑的现有模型与预测编码模型相结合来克服这一障碍 头脑和大脑。将预测编码融入焦虑模型将提供更好的理解 神经活动如何与对焦虑很重要的主观体验相关(例如，威胁感知)。该项目 测试预测编码提出的关于威胁感知神经表征的两个平行假设 模型：威胁感知的神经表示是特定于内容的(目标1)，并且神经表示 威胁感知的表现取决于预期(目标2)。使用单一的设计，我们操纵 内容-特异性和期望值，以平行检验这两个假设。我们使用功能磁共振成像来测量大脑 活动，并使用自我报告和外围心理生理学来测量主观的威胁体验 感知力。受试者发现刺激具有威胁性的程度(根据自我报告和 心理生理学)，我们假设我们将观察到相对特定于内容的神经表征 威胁。我们还假设，在预期和非预期条件下，威胁的神经表征将有所不同。 违反预期。预期的影响可能会影响特定内容的神经活动或核心活动 一组区域。从拟议的项目中获得的知识有可能提高对 神经活动和主观体验之间的映射。与翻译神经科学相关，更好的 了解焦虑的心理和神经生物学机制将是缩小差距的关键 在实验室研究和更有效的焦虑症治疗之间。更广泛地说，预测编码模型 是基本大脑功能的模型。因此，我们提出的预测编码模型提供了一种新的理论 可概括到涉及无序威胁感知的精神疾病的框架(例如， 精神分裂症)和其他情感障碍(如抑郁症、双相情感障碍)。

PROJECT SUMMARY Anxiety disorders affect an estimated 6.7% to 7.3% of people globally each year and incur a large burden on people’s lives. To address this urgent problem, it is necessary to develop a better understanding of the neural bases of subjective experiences in anxiety, including threat perception. Translational neuroscience has focused on animal models of defensive behavior involving a core set of regions. Although these animal models and the human subjects research they inspired have led to advances in treatment of anxiety, the mapping between neural mechanisms and subjective experience remains poorly understood. The set of regions found to support defensive behavior in animal models does not appear to be involved in all instances of fear or anxiety. The current project overcomes this barrier by integrating current models of anxiety with predictive coding models of the mind and brain. Incorporating predictive coding into models of anxiety will offer a better understanding of how neural activity relates to subjective experiences important to anxiety (e.g., threat perception). The project tests two parallel hypotheses about neural representation of threat perception suggested by predictive coding models: that neural representations of threat perception are content-specific (Aim 1) and that neural representations of threat perception depend on expectations (Aim 2). Using a single design, we manipulate content-specificity and expectations to test these two hypotheses in parallel. We use fMRI to measure brain activity and use self-report and peripheral psychophysiology to measure subjective experiences of threat perception. To the extent that participants find stimuli threatening (as indexed by self-report and psychophysiology), we hypothesize that we will observe relatively content-specific neural representations of threat. We also hypothesize that neural representations of threat will differ under conditions of expectation vs. expectancy violation. The effect of expectation may impact content-specific neural activity or activity in a core set of regions. The knowledge gained from the proposed project has the potential improve understanding of the mapping between neural activity and subjective experiences. Relevant for translational neuroscience, a better understanding of the psychological and neurobiological mechanisms of anxiety will be critical to closing the gap between laboratory research and more effective treatments for anxiety. More broadly, predictive coding models are models of basic brain function. Thus, the predictive coding model we propose offers a new theoretical framework that is generalizable to psychiatric illnesses involving disordered threat perception (e.g., schizophrenia) and other affective disorders (e.g., depression, bipolar disorder).