Computational and neurodevelopmental mechanisms of memory-guided decision-making

记忆引导决策的计算和神经发育机制

• 批准号: 10723314
• 负责人: Catherine Insel
• 金额: $ 13.08万
• 依托单位: COLUMBIA UNIV NEW YORK MORNINGSIDE
• 依托单位国家: 美国
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-07-03 至 2025-06-30
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10723314
• 关键词: Acceleration; Adaptive Behaviors; Adolescence; Adolescent; Adult; Age; Award; Behavior; Behavioral; Categories; Clinical; Computer Models; Corpus striatum structure; Data; Data Set; Decision Making; Development; Eating; Episodic memory; Event; Failure; Food; Fostering; Foundations; Functional Magnetic Resonance Imaging; Goals; Hippocampus; Individual; Knowledge; Learning; Link; Longevity; Machine Learning; Measures; Mediating; Memory; Mental Depression; Mental Health; Methods; Modeling; Outcome; Pathway interactions; Phase; Postdoctoral Fellow; Prefrontal Cortex; Process; Psychopathology; Research; Research Training; Rewards; Risk; Sampling; Sesame - dietary; Shapes; Specific qualifier value; Specificity; Support System; System; Techniques; Testing; Theoretical model; Training; Work; Youth; adolescent brain development; artificial neural network; career; cognitive neuroscience; cognitive process; cognitive testing; computational neuroscience; data de-identification; depressive symptoms; experience; longitudinal design; maladaptive behavior; neural; neural network; neurobiological mechanism; neurodevelopment; neuroimaging; novel; predictive modeling; research study; risk prediction; translational approach; two-dimensional

PROJECT SUMMARY Adolescence is accompanied by enhanced risk for maladaptive decision-making, a profile that confers risk for psychopathology. Prior work has shown that memory and decision-making systems continue to mature during adolescence, however, developmental studies have traditionally tested these cognitive processes in isolation. As a result, remarkably little is known about how the development of memory and the corresponding neural processes influence adaptive and maladaptive behaviors in youth. The proposed project will test a novel conceptual model that specifies the neurodevelopmental pathways through which memory guides decision- making during adolescence. This unifying framework formalizes the neurocomputational mechanisms underlying two dimensions of memory-guided decision making: specificity and generalization. The proposed research will test this conceptual model by leveraging recent advances from computational neuroscience, including artificial neural networks, multivariate neuroimaging methods, and machine learning approaches to predictive modeling. Aim 1 (K99 phase) tests the hypothesis that enhanced interactions between the hippocampus and striatum support the use of memory specificity to guide decisions. Aim 2 (K99 phase) tests the hypothesis that enhanced interactions between the hippocampus and vmPFC support the use of memory generalization to guide decisions. Aim 3 (R00 phase) tests the overarching hypothesis that adolescents rely on specificity to guide decision-making due to the early maturation of hippocampal-striatal circuitry, whereas generalization emerges during the transition to adulthood, in tandem with strengthening connections between the hippocampus and ventromedial prefrontal cortex (vmPFC). Further, Aim 3 will test the hypothesis that decreased hippocampal-striatal interactions predict risk for depressive symptoms during adolescence. This work directly builds on the candidate’s earlier postdoctoral work that has characterized the behavioral trajectories of memory-guided decision-making during adolescence. By examining the computational foundations and neurodevelopmental trajectories of memory-guided decision-making, the proposed research will reveal how learning and decision- making processes refine with age during adolescence. This work has the potential to illuminate why adolescence is a period of heightened risk for maladaptive decision-making, and the proposed research will identify how aberrant decision-making confers risk for depression. This award will provide the candidate, who has a strong background in developmental cognitive neuroscience, with critical training in computational modeling with neural networks, multivariate fMRI methods, machine learning, and advanced statistical techniques to promote a successful transition to an independent research career. The proposed research study and training plan will help the candidate achieve her ultimate goal of leading an independent research lab, where she will implement computational and translational approaches to study the neurodevelopment of adaptive and maladaptive behaviors during adolescence.

项目摘要 青春期伴随着适应不良决策的风险增加，这一特征赋予了风险， 精神病理学先前的研究表明，记忆和决策系统在发育过程中不断成熟。 然而，青春期的发展研究传统上是孤立地测试这些认知过程。 因此，关于记忆的发展和相应的神经系统是如何发展的， 过程影响青年的适应和适应不良行为。拟议中的项目将测试一部小说 概念模型，它指定了记忆引导决策的神经发育途径， 在青春期做的。这个统一的框架形式化的神经计算机制， 记忆引导决策的两个维度：特异性和概括性。拟议的研究将 通过利用计算神经科学的最新进展来测试这个概念模型， 神经网络、多变量神经成像方法和用于预测建模的机器学习方法。 目的1（K99期）验证海马和纹状体之间增强的相互作用的假设 支持使用记忆特异性来指导决策。目的2（K99阶段）测试增强的假设 海马体和VMPFC之间的相互作用支持使用记忆概括来指导决策。 目标3（R 00阶段）检验青少年依赖特异性来指导决策的总体假设 这是由于海马-纹状体回路的早期成熟，而泛化出现在 过渡到成年，与加强海马体和腹内侧之间的连接 前额叶皮层（vmPFC）。此外，目标3将检验减少的海马-纹状体 交互作用预测青春期抑郁症状的风险。这项工作直接建立在 候选人早期的博士后工作，其特点是记忆引导的行为轨迹 青春期的决策。通过检查计算基础和神经发育 记忆引导决策的轨迹，拟议的研究将揭示如何学习和决策， 在青少年时期，随着年龄的增长，这些过程变得更加精细。这项工作有可能阐明为什么青春期 是适应不良决策风险增加的时期，拟议的研究将确定如何 异常的决策会导致抑郁症。这个奖项将提供候选人，谁有一个强大的 有发展认知神经科学的背景，在神经网络计算建模方面受过严格的培训。 网络，多变量功能磁共振成像方法，机器学习和先进的统计技术，以促进一个 成功过渡到独立研究生涯。拟议的研究和培训计划将有助于 候选人实现了领导一个独立研究实验室的最终目标，在那里她将实施 计算和翻译的方法来研究适应性和适应不良的神经发育 青春期的行为。