Neurocomputational Mechanisms of Affective Semantic Memory Development

项目总结/摘要 我的职业目标是领导一个跨学科的团队，研究发展神经计算 将早期生活逆境与心理健康联系起来的机制。为了达到这个目标，我需要在以下方面进行强化训练： 人类神经成像方法、认知计算神经科学和二元方法。我的训练 到目前为止，已经提供了一个强大的基础技能，在发展神经科学，早期生活的逆境，统计 建模、经验采样和机器学习。我找到了最好的导师团队， 科学的环境来支持我扩展这一技能，在情感记忆方面进行必要的训练， 功能磁共振成像分析，计算建模，现实世界的二元（亲子）行为方法，和专业 skills.有了这些受保护的培训时间，我将准备好成功地过渡到一个有竞争力的博士后 位置，并最终获得在发展神经科学的前沿终身教职。 研究：超过三分之一的儿童经历早期的逆境，如虐待，忽视， 父母遗弃/分离，至少占30 - 45%的精神健康障碍的发病原因 世界范围内。临床关联研究表明，这种精神病理学风险的增加与以下因素有关： 中线皮质-皮质下回路的扰动，包括中扣带-岛叶"显着"网络， 后内侧"默认模式"网络。同时，认知神经科学的实验研究发现， 这条线路在语义、社会和情感知识中具有更广泛的功能作用。神经系统 在早期经历逆境后引起精神病理学风险的变化是同一种， 神经生物学发现代表学习的情感语义记忆？根据初步数据，我 将检验这一假设，即早期逆境后中线皮质-皮质下回路的改变代表了 在早期学习经历中习得的情感语义知识。在F99阶段，我将 确定与儿童情感语义记忆相关的中线皮质-皮质下活动模式 在功能磁共振成像过程中，在发育样本中富集早期发育逆境（目标1.1），然后检查 情感语义神经表征与现实世界亲子情感行为之间的联系， 二元生态瞬时评估（目标1.2）。在K00阶段，我将确定贝叶斯 中线皮质-皮质下回路中的预测编码解释了情感的神经计算 儿童时期的语义记忆，并研究这些神经计算机制如何不同的功能， 年龄、社会情绪背景和早期逆境暴露（目标2）。这项研究将产生更强大的， 生态敏感的人类发展计算模型，通过整合尖端的 神经计算方法，实时亲子行为和情感语义记忆。等 这些进展有可能显著扩展精神病理学的神经发育模型， 对于开发针对年龄，早期经验和神经计算机制的干预措施至关重要。

PROJECT SUMMARY / ABSTRACT My career goal is to lead an interdisciplinary team that investigates the developmental neurocomputational mechanisms that link early-life adversity to mental health. To reach this goal, I require intensive training in human neuroimaging methods, cognitive computational neuroscience, and dyadic methodologies. My training to date has provided a strong foundation of skills in developmental neuroscience, early-life adversity, statistical modeling, experience sampling, and machine learning. I identified the best possible mentorship team and scientific environment to support me in expanding this skill set with essential training in emotional memory, fMRI analysis, computational modeling, real-world dyadic (parent-child) behavior methods, and professional skills. With this protected training time, I will be poised to successfully transition to a competitive postdoctoral position and ultimately obtain a tenure-track faculty position at the forefront of developmental neuroscience. Research: More than one-third of children experience early caregiving adversity such as abuse, neglect, and parental abandonment/separation, which accounts for the onset of at least 30-45% of mental health disorders world-wide. Clinical association studies show that this increased risk for psychopathology is linked to perturbations in midline cortico-subcortical circuitry, comprised of the midcingulo-insular “salience” network and posterior-medial “default mode” network. At the same time, experimental studies in cognitive neuroscience find that this circuitry has a broader functional role in semantic social and affective knowledge. Might the neural alterations that give rise to psychopathology risk following early caregiving adversity be one and the same as the neurobiology found to represent learned affective semantic memories? In line with my preliminary data, I will test the hypothesis that alterations in midline cortico-subcortical circuitry following early adversity represent the affective semantic knowledge learned during early caregiving experiences. During the F99-phase, I will identify the midline cortico-subcortical activity patterns associated with children’s affective semantic memories during fMRI in a developmental sample enriched for early caregiving adversity (Aim 1.1), and then examine the links between affective semantic neural representations and real-world parent-child emotional behaviors using dyadic ecological momentary assessments (Aim 1.2). In the K00-phase, I will determine whether Bayesian predictive coding in midline cortico-subcortical circuitry explains the neural computations underlying affective semantic memory in childhood and examine how these neurocomputational mechanisms differ as a function of age, socioemotional context, and early adversity exposure (Aim 2). This research will result in more powerful, ecologically-sensitive computational models of human development through the integration of cutting-edge approaches to neurocomputation, real-time parent-child behavior, and affective semantic memory. Such advances have the potential to significantly extend neurodevelopmental models of psychopathology, which is essential for developing interventions tailored to age, early experience, and neurocomputational mechanism.