A Probabilistic Concept of Sensory Cortical Funcion

感觉皮层功能的概率概念

• 批准号: 6749068
• 负责人: DALE PURVES
• 金额: $ 18.29万
• 依托单位: DUKE UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2003
• 资助国家: 美国
• 起止时间: 2003-07-01 至 2006-06-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/6749068
• 关键词: Scandentia; behavioral /social science research tag; brain mapping; form /pattern perception; image processing; neural information processing; orientation; sensory mechanism; space perception; statistics /biometry; stereotaxic techniques; visual cortex; visual perception; visual stimulus

DESCRIPTION (provided by applicant): Recent work has led to the hypothesis that the percepts generated by visual cortex are determined stochastically according to the success (or failure) of past behaviors in response to inevitably ambiguous retinal images. So far, the basis for this idea is indirect: psychophysical evidence shows that the images seen in response to a wide variety of stimuli accord with the probability distributions of the possible sources of the corresponding retinal image. These findings imply that the primary role of the visual cortical processing -- and by inference the role of sensory cortices generally -- is to instantiate and process the probability distributions of the possible stimulus sources. Here we propose to initiate a direct exploration of this novel concept of cortical function by taking advantage of the ease (and accuracy) with which visuotopic and orientation maps can be determined in the primary visual cortex (V1) by optically imaging patterns of cortical activity. Using this method, we propose to examine how geometrical target angles that elicit very different perceptions as a function of context are represented in an experimental animal (the tree shrew to begin with). The project entails three Specific Aims: 1) To determine with optical imaging of 'intrinsic signal' how simple angular stimuli are represented in VI of anesthetized animals, using both topographic and orientation maps as indices; 2) To determine the cortical representations of topography and orientation elicited by the same stimuli in contexts that markedly change the probability distributions of the possible sources (and thus perception); and 3) To examine whether awake behaving animals, much as human subjects, see the same geometrical stimuli differently as a function of contextual statistics. These experiments will show whether visuotopic and orientation maps represent image features (i.e., location in space and line orientation), as has long been thought, or whether, as we suspect, they represent the probability distributions of the possible sources of visual stimuli. Knowledge that the primary visual cortex operates in this probabilistic way would demand a reinterpretation of present structural and functional information about V1, and would strongly encourage further exploration of other sensory cortices using this framework.

描述（由申请人提供）： 最近的工作导致了以下假设：视觉皮层产生的感知是根据过去行为的成​​功（或失败）对不可避免地模棱两可的视网膜图像而随机确定的。到目前为止，这一想法的基础是间接的：心理物理证据表明，对各种刺激的响应图像符合相应视网膜图像的可能来源的概率分布。这些发现表明，视觉皮质加工的主要作用 - 以及通过推断，感觉皮质的作用通常 - 是实例化和处理可能刺激源的概率分布。在这里，我们建议通过利用轻松（和准确性）来直接探索这种新型皮质功能的概念，在这些概念中，可以通过光学成像的皮质活动模式在主要视觉皮层（V1）中确定视觉和方向图。使用这种方法，我们建议研究如何在实验动物中表示引起截然不同的看法作为上下文的函数的几何目标角度（从开始的树木开始）。该项目需要三个特定的目的：1）用“固有信号”的光学成像确定简单的角度刺激在麻醉动物的VI中如何使用地形图和方向图作为指数表示； 2）确定在明显改变可能来源的概率分布（从而感知）的情况下，相同刺激引起的地形和方向的皮质表示； 3）检查醒着的动物是否像人类受试者一样，将相同的几何刺激视为上下文统计的函数。这些实验将表明视觉和方向图映射是否代表图像特征（即，在空间和线方向中的位置），或者长期以来一直认为是图像特征，还是正如我们所怀疑的那样，它们是否代表了可能的视觉刺激来源的概率分布。知道主要视觉皮层以这种概率方式运行的知识将要求重新解释有关V1的当前结构和功能信息，并强烈鼓励使用此框架进一步探索其他感觉皮质。