Cortical Interactions Underlying Sensory Representations

项目摘要 感官知觉包括处理传入的感官输入并通过规则解释该信息 是由先前的经验产生的。刺激特征需要结合在一起，以形成更复杂的感觉 然后与效价或动作结果相关联，以赋予这些表示以意义。 在哺乳动物的新皮层中，感觉表征的形成被认为是通过 分布在几个皮层区域的处理过程。在这个框架之外， 将感官信息转化为越来越抽象的 代表性仍然未知。要实现更深层次的机制性理解， 理论模型和实验工作之间的循环，并通过 跨模式系统的比较方法。许多主要的感官理论模型 加工来源于在人类或非人类灵长类动物中进行的实验研究。它 由于有限的实验数据，很难验证或改进这些计算模型。 在这些物种中获得用于回路级解剖的工具。然而，用于电路解剖的新工具是 现在可用于小鼠。使用新开发的基于胡须的鼠标行为， 概括灵长类动物的知觉任务，我们建议调查刺激信息是如何编码的 并在初级躯体感觉、次级躯体感觉和嗅周皮层之间转换， 在感觉处理的日益复杂的阶段中起作用的突出的神经元连接区域。我们 我们的目标是实现电路级的理解如何刺激必然会产生高阶 表征，这些表征如何与行动结果相关联，以及它们是如何学习的 并根据行为要求召回。这项工作将提供关键的新的见解如何本地和远程 皮层回路的功能是建立内部表征和评估外部环境。通过测试和 在改进哺乳动物物种的皮质功能模型的基础上，我们将寻求得出共同的原则， 电路功能，解释如何越来越不变和抽象的表示可能被编码， 在大脑皮层中实现。

PROJECT SUMMARY Sensory perception involves processing incoming sensory input and interpreting that information through rules generated from prior experience. Stimulus features need to be bound together to form more complex sensory representations and then associated with a valence or action outcome to give meaning to those representations. In the mammalian neocortex, the formation of sensory representations is believed to occur through processing that is distributed across several cortical areas. Beyond this general framework, the exact circuits and computations involved in transforming sensory information into increasingly abstract representations remain unknown. To achieve a deeper mechanistic understanding, it is necessary to close the loop between theoretical models and experimental work and to identify common mechanisms through comparative approaches across model systems. Many of the leading theoretical models for sensory processing have been derived from experimental studies performed in humans or non-human primates. It has been difficult to validate or refine these computational models because of the limited experimental access to tools for circuit-level dissection in those species. However, novel tools for circuit dissection are now available for applications in mice. Using newly developed whisker-based mouse behaviors that recapitulate perceptual tasks in primates, we propose to investigate how stimulus information is encoded and transformed across primary somatosensory, secondary somatosensory, and perirhinal cortex, three prominent reciprocally connected areas that function at increasingly complex stages of sensory processing. Our goal is to achieve a circuit-level understanding for how stimuli are bound to generate higher-order representations, how such representations are associated with action outcomes, and how they are learned and recalled as required for behavior. This work will provide critical new insight into how local and long-range cortical circuits function to build internal representations and evaluate the external environment. By testing and improving upon models of cortical function across mammalian species, we will seek to derive common principles of circuit function that explain how increasingly invariant and abstract representations may be encoded and implemented in the neocortex.