Computational Analysis of Hippocampal Function

认知神经科学的一个主要目标是根据神经回路的生理特性来理解行为和心理过程。海马区的定位细胞是研究这些问题的理想系统。这些电池发射强劲，特别是在环境中的有限位置。它们被认为编码了一个空间框架，用于组织经验的项目和事件，形成了情景记忆和灵活的、上下文相关的学习的基础。研究这些细胞的一个重大进展是开发了记录数十个分离良好的神经元活动的技术 同时从海马体结构的不同部分发出。这些技术为实验者提供了测试有影响力的海马体功能计算理论的工具，这些理论已经存在了几十年，而没有在单一单位生理学水平上进行严格的测试。 我们建议开发一个在海马结构中处理的计算模型，以帮助分析和解释我们实验室的集合单元记录。目标是了解将每个级别的输入属性转换为输出属性的机制，并从这些机制推断功能角色。为此目的，将有必要对反应性质进行计算机模拟，以便对不那么正式的分析不会产生的数据有深入的了解；对原本不会设想的数据进行更多的分析；并激发新的实验，以解决计算分析产生的问题和假设。 这些分析将考察(1)全局抑制机制在位置场重新映射中的作用，(2)重复的侧支回路在位置场对受控环境操作的动态反应中的作用，以及(3)神经可塑性，特别是依赖于峰时的可塑性，在位置细胞对受控环境操作的动态反应中的作用。阿尔茨海默病和癫痫等疾病的破坏性神经影响与海马体和相关脑区的功能障碍密切相关。这些研究将产生对这些构成学习和记忆基础的大脑区域之间的神经相互作用的基本见解，以及对这些令人衰弱的疾病中这些机制如何出错的见解。

A major goal of cognitive neuroscience is to understand behavior and mental processes in terms of the physiological properties of neural circuits. Place cells of the hippocampus are an ideal system to study these issues. These cells fire robustly and specifically in restricted locations in an environment. They are thought to encode a spatial framework used to organize the items and events of experience, forming the foundation of episodic memory and flexible, context-dependent learning. A major advance in the study of these cells has been the development of techniques to record the activity of dozens of well-isolated neurons simultaneously from different parts of the hippocampal formation. These techniques provide experimenters the tools to test influential computational theories of hippocampal function that have existed for decades without rigorous testing at the level of single-unit physiology. We propose to develop a computational model of processing in the hippocampal formation to aid in the analysis and interpretation of ensemble unit recording from our laboratory. The goal is to understand the mechanisms that transform the input properties at each level into the output properties, and to deduce functional roles from these mechanisms. To this end, it will be necessary to implement computer simulations of the response properties, in order to engender insights into the data that would not arise from less formal analysis; to stimulate additional analysis of the data that would not have otherwise been conceived; and to inspire new experiments to address issues and hypotheses that result from the computational analyses. These analyses will investigate (1) the role of global inhibitory mechanisms in the remapping of place fields, (2) the role of recurrent collateral circuitry in the dynamic responses of place fields to controlled environmental manipulations, and (3) the role of neural plasticity, in particular spike-timing dependent plasticity, in the dynamic responses of place cells to controlled environmental manipulations. The devastating neurological effects of such diseases as Alzheimer's Disease and epilepsy are intimately tied to dysfunctions of the hippocampus and related brain areas. These studies will generate fundamental insights into the neural interactions between these brain areas that underlie learning and memory, as well as insights into how these mechanisms go awry in these debilitating diseases.