Imaging Mechanisms of Action in Motor Learning

在美国国家科学基金会的支持下，熊博士将开发成像和建模策略来研究人类大脑适应性变化的潜在机制。本研究的重点是探讨运动学习的机制。学习诱导的神经可塑性和功能重组已经得到了很好的证明，但还没有得到很好的理解。目前的神经影像学研究通过探索区域神经活动和任务表现的区域间活动的变化来研究学习的神经机制。很少有人研究神经连接和突触权重的更基本的变化。在技术方面，人体功能成像研究迫切需要更严格的方法，这可以通过数学建模来提供。一个建模框架-结构方程建模-现在被认为是适合于人体成像数据。然而，结构方程建模目前是基于非人类物种神经解剖学研究的解剖约束进行的。利用同一受试者的任务无关的解剖连通性数据对解剖约束进行单独优化，可以大大提高结构方程建模的性能。迄今为止，还没有任何实验室报告过这种策略。本研究旨在开发神经成像的系统级建模策略，并将这些新策略应用于运动学习的作用机制。本提案的总体目标将通过以下四个目标来实现。首先，开发和优化成像策略，以检测每个个体的解剖连通性。其次，通过结合个体解剖约束来开发结构方程建模策略，以提高这些模型的性能。第三，利用增强建模策略研究运动学习诱导的区域神经活动和区域间活动的变化。第四，应用增强模型研究突触可塑性，证明突触可塑性是运动学习的潜在作用机制。完成后，该研究项目将增加对适应性学习机制的理解，并有可能定义一种新的策略，通过该策略，功能成像可以应用于研究作用机制和疾病病理生理学。

With National Science Foundation support, Dr. Xiong will develop imaging and modeling strategies to study mechanisms underlying adaptive changes of the human brain. The focus of this proposal is to explore the mechanisms underlying motor learning. Learning-induced neural plasticity and functional reorganization are well-established and well-documented, but not well-understood. Current neuroimaging studies investigate neural mechanisms underlying learning by exploring the changes in regional neural activity and inter-regional activity of task-performance. Little effort has been given to studying the more fundamental changes of neural connections and synaptic weighting. On the technical front, human functional imaging research sorely needs more rigorous approaches, as can be provided by mathematical modeling. A modeling framework - Structural Equation Modeling - is now accepted as appropriate for human imaging data. Structural equation modeling however, is currently performed with anatomical constraints based on neuroanatomical studies in non-human species. The performance of structural equation modeling might be greatly enhanced if anatomical constraints are individually optimized using the same subject's task-independent anatomical connectivity data. To date, this strategy has not been reported by any laboratory. The present proposal seeks to develop system-level modeling strategies for neuroimaging and to apply these novel strategies to mechanisms of action of motor learning. The overall goal of this proposal will be accomplished through the following four goals. First, developing and optimizing imaging strategies for detecting anatomical connectivity for each individual subject. Second, developing a structural equation modeling strategy by incorporating individual anatomical constraints to enhance those models' performance. Third, investigating changes in regional neural activity and inter-regional activity of task-performance induced by motor learning using the enhanced modeling strategy. Fourth, investigating synaptic plasticity by applying the enhanced modeling and demonstrating that synaptic plasticity is an underlying mechanism of action of motor learning. When completed, this research project will increase the understanding of mechanisms of adaptive learning and has the potential of defining a new strategy by which functional imaging can be applied to study mechanisms of action and disease pathophysiology.