Dynamic Visual Activity in Temporal Cortex

描述：神经系统如何从我们这个复杂而动态的世界中提取出连贯的观点，目前还没有得到很好的理解。目前对涉及视觉物体识别的神经回路的理解，在很大程度上是基于被动动物观看静态图像时所做的生理记录。现实世界的视觉与这些条件有很大的不同，因为观察者和世界中的物体都是动态的。此外，主动视觉是有目的的；观察者的感知可以引导可能反过来改变视觉环境的行动。本实验的目的是通过记录猴子在长时间观察、监测和互动复杂视觉形式时，颞叶外皮层视觉区产生的神经活动，开始研究主动视觉的神经基础。第一个具体目标将探讨如何知道什么时候会发生的事情准备感知系统，以确定发生了什么。第二个目标将测试一个假设，即IT中的神经活动不仅由物体的可见存在维持，而且当物体暂时从视野中消失时，在与它的持续物理存在一致的环境中也能维持。最终目的是考虑是否针对一个物体的公开行为，或者对即将发生的事件的计划行为，直接影响IT神经元对视觉刺激的反应。现有的客体处理神经生理学理论认为，颞叶皮层中的细胞对于将获得的视觉快照与存储的表征相匹配至关重要，但这些大脑区域如何促进持续的视觉体验，这一问题尚未解决。通过对细胞的记录，其假定的作用是传递有关详细世界的信息，预计这些实验可以提供直接解决这一问题的基本经验数据，并且这样做将迫使考虑动态神经处理理论，而不是传统的视觉感知“快照”模型。

(Revised Abstract) Description: How the nervous system extracts a coherent view of our complex and dynamic world is not well understood. Current understanding of the neural circuitry involved in visual object recognition is based in large part on physiological recordings made during passive animals looking at static images. Real world vision differs dramatically from these conditions, as both the observer and objects in the world are dynamic. Further, active vision is purposive; an observer's percepts can guide actions that may in turn alter the visual environment. The objective of the experiments described in this proposal is to begin to investigate the neural basis of active vision by recording the neural activity generated in extrastriate visual areas of the temporal lobe while monkeys observe, monitor, and interact with complex visual forms over temporally extended periods.The first specific aim will explore how knowing when something will occur prepares the perceptual system to determine what has occured. The second aim will test the hypothesis that neural activity in IT is maintained not only by the visible presence of the object, but also when that object disappears momentarily from view but in a context consistent with its continued physical presence. The final aim considers whether either overt action directed toward an object, or planned action for an upcoming event, directly affect the responses of IT neurons to visual stimuli. Existing neurophysiological theories of object processing propose that cells in the temporal cortex are essential for matching acquired visual snapshots to stored representations, but they leave open the question of how these brain areas contribute to continuous visual experience. By recording from cells whose putative role is to convey information about the detailed world, it is expected that these experiments can provide essential empirical data directly addressing this issue, and in doing so, will force the consideration of theories of dynamic neural processing as opposed to traditional "snapshot" models of visual perception.