Training-related motor plasticity and the representation of skill

运动技能是我们与世界互动的一个重要组成部分，然而我们的大脑结构和功能如何支持他们获得的模型严重不足-几乎完全缺乏支持神经可塑性的生理机制的任何证据。大多数运动神经可塑性研究只使用一组参与者在短时间内（主要是在一天或两天内）接受单一任务的训练，这使得很难确定与学习本身相关的过程，也很难得出自然学习的结论，因为自然学习通常需要几天、几周或几个月的训练。此外，成像研究绝大多数只研究一两个成像指标，这些指标既不是定量的，也不是与脑生理学密切相关的；这使得进行纵向比较变得困难，也不可能对可塑性的生理标记做出推断。如果没有一个更完整的生理过程（如髓鞘形成、树突状乔木化和血管增生）如何支持可塑性的模型，我们的理解将局限于“变化”发生的地点和时间，没有迹象表明发生了什么变化。因此，我们必须发展一个更有生理学知识的框架，以产生人类神经可塑性时间进程的更具体的机制模型。我的总体研究目标是更深入地了解大脑如何通过神经可塑性过程支持运动技能的获得。本研究的短期目标是通过以下方式建立一个基于生理学的人类运动神经可塑性模型：确定大脑中支持序列特异性可塑性本身的区域（超越表现的影响），描述其变化的时间过程，并确定支持神经可塑性的假定神经生理学机制。所提出的工作是基于一个独特而丰富的超高场（7T）多模态定量磁共振成像（MRI）数据集，该数据集之前由我和合作者收集。两组20名参与者（主动组和对照组）接受连续5天的精细运动控制视觉-运动序列学习任务（顺序捏力任务）训练。每天获得多模态定量MRI，包括对髓磷脂、细胞密度、组织微观结构、静息血流量、血容量和铁敏感的序列。这个独特的数据集将被HQP（包括2名博士和3名硕士）用来开发一个结构、功能和多模态训练相关的神经可塑性生理学模型，关键是利用我们收集的多模态定量MRI指标的生理起源知识来确定与训练相关的神经可塑性的机制时间过程。总之，拟议的研究是对我们的大脑如何学习运动技能的机制理解的关键的第一步。

Motor skills are a crucial component of our interaction with the world, yet our models for how the structure and function of the brain support their acquisition are severely underspecified - almost completely lacking any evidence for the physiological mechanisms that support neuroplasticity. Most motor neuroplasticity studies use only a single group of participants trained on a single task over a short period of time (predominantly within a single day or across two), making it difficult to identify processes linked to learning per-se or draw conclusions about naturalistic learning that often requires days, weeks, or months of training. In addition, imaging studies overwhelmingly investigate only one or two imaging metrics that are neither quantitative nor strongly linked to brain physiology; making it both difficult to perform longitudinal comparisons and impossible to make inferences about the physiological markers of plasticity. Without a more complete model for how physiological processes (such as myelination, dendritic arborization, and vascular proliferation) support plasticity, our understanding will be limited to where and when "change" occurred, with no indication of what changed. Thus, we must develop a more physiologically-informed framework to generate more specific mechanistic models of the timecourse of human neuroplasticity. My overall research goal is to develop a deeper understanding of how the brain supports the acquisition of motor skills through the process of neuroplasticity. The short-term objectives porposed here will develop a physiologically-informed model of human motor neuroplasticity by: determining the regions of the brain that support sequence-specific plasticity per-se (over and above the effect of performance), characterizing their timecourses of change, and identifying the putative neurophysiological mechanisms that support neuroplasticity. The proposed work is based on a unique and rich Ultra-high field (7T) multimodal quantitative magnetic resonance imaging (MRI) dataset that was previously collected by myself and colloborators. Two groups of 20 participants (active, control) were trained on a visuo-motor sequence learning task of fine motor control (sequential pinch force task) over 5 consecutive days. Multimodal quantitative MRI was acquired on each day, and included sequences sensitive to myelin, cell density, tissue microstructure, resting blood flow, blood volume, and iron. This unique dataset will used by HQP including 2 PhDs and 3 Masters to develop a physiologically-informed model of structural, functional and multimodal training-related neuroplasticity, crucially leveraging our knowledge of the physiological origins of the collected multimodal quantitative MRI metrics to identify the mechanistic timecourse of training-related neuroplasticity. Together, the proposed research is a critical first step towards a mechanistic understanding of how our brains learn motor skills.