NEURAL BASIS OF INTERNAL REPRESENTATION OF PLACE

描述（改编自申请人摘要）： 空间关系为大多数适应性行为提供了背景， 情景记忆的框架 他们的内部代表涉及一个 从原始感觉输入的自我中心坐标转换为 非中心的框架，并最终转化为运动输出， 以目标的空间坐标表示。 机制 空间编码在哺乳动物中是高度保守，他们的研究提供了 深入了解高级认知过程的神经基础和因素， 导致早期发育、衰老、脑损伤中的功能异常， 疾病和药物滥用。 这项研究是由一个不断发展的理论指导的 其中地标和事件被映射到二维度量上 框架或“图表”，用于连接大脑的表征 结构. 它们的坐标是由一群 海马“定位”细胞。 该指标基于网络架构 通过线性和角度的自我运动来连接图表坐标 信息（“路径整合”）。 图表在突触中预先配置 连接，独立于外部输入，但它们成为关联的 通过探索获得地标信息。 这使得适当的 框架的选择，校正漂移误差，这是固有的任何 路径集成系统，并使用高效的矢量类操作， 计算到目标的轨迹，尽管后面的计算可能 不是在海马体中进行，而是在其新皮层靶区中进行。 理论 是通过同时记录大量的神经元 （50-150）和空间的神经群体代码的解释 经验和行为。 拟议的技术发展应 充分增加这些数字，以便准确阅读人口代码 研究计算机制所需的短时间尺度上， 多个大脑区域之间的神经系统相互作用。 当前 研究问题是：1）环境结构和空间 海马活动编码的事件背景？ 2)要么 导航目标或最近访问过的位置的记忆 在海马体或新皮层中表示，并且前者编码在一个 这种方式可以帮助最短路线的矢量式计算？ 第三章 什么是作用的筋膜齿，一个主要的海马子区， 空间学习所需，但不适用于特定位置的发射。 海马锥体细胞？ 4)是海马可塑性机制， 可能是空间学习的基础， interneuron放电that occur发生in novel新environments环境？ 数据将 考虑到有关的数值模拟提出的机制 使用简化的尖峰神经元模型。

DESCRIPTION (Adapted from applicant's abstract) : Spatial relationships provide the context for most adaptive behaviors and the framework for episodic memory. Their internal representation involves a transformation from the egocentric coordinates of raw sensory inputs into an allocentric framework, and ultimately into motor outputs that are also represented in terms of the spatial coordinates of the goal. Mechanisms of spatial coding are highly conserved across mammals, and their study provides insight into the neural basis of higher cognitive processes and factors that contribute to abnormal function in early development, aging, brain injury, disease, and substance abuse. The research is guided by an evolving theory in which landmarks and events are mapped onto 2-dimensional metric frameworks or "charts" that are used to connect representations of brain structures. Their coordinates are signaled by the activity of groups of hippocampal "place" cells. The metric is based on a network architecture that links chart coordinates through linear and angular self-motion information ("path-integration"). Charts are preconfigured in the synaptic connections, independently of external input, but they become associated with landmark information through exploration. This enables appropriate framework selection, correction for the drift error that is inherent in any path integration system and the use of efficient, vector-like operations to compute trajectories to goals, although the latter computations may be carried out not in hippocampus, but in its neocortical targets. The theory is explored using simultaneous recordings from large groups of neurons (50-150) and the interpretation of neural population codes for spatial experiences and behaviors. Proposed technological developments should increase these numbers sufficiently for accurate reading of population codes on the short time-scales necessary to study computational mechanisms and the interaction of neural ensembles between multiple brain regions. The current research questions are: 1) How are environmental structure and the spatial context of events encoded in hippocampal activity? 2) Are either navigational goals or memories of recently visited locations explicitly represented in the hippocampus or neocortex and are the former encoded in a manner that could subserve vector-like computations of shortest routes? 3) What is the role of the fascia dentata, a major hippocampal subfield that is required for spatial learning, but not for place specific firing in hippocampal pyramidal cells? 4) Are hippocampal plasticity mechanisms that presumably underlie spatial learning modulated by the suppression of interneuron firing that occurs in novel environments? The data will be considered in relation to numerical simulations of the proposed mechanisms using simplified, spiking-neuron models.