Predictive Representation of Motion in Visual Cortex

描述（由申请人提供）：大量证据表明，视觉感知受到我们对视觉世界的预测的影响。预测使我们能够利用和适应视觉环境的因果、非随机结构，对于快速、准确地控制视觉引导行为至关重要。我们之前在猴子的外侧顶叶内区（LIP）中发现了神经元信号，这些信号提供了视觉运动方向的预测表示。在这个新的建议中，首要的问题是预测信号与行为的关系以及预测信号发展的机制。LIP中的预测运动信号似乎非常适合行为的前馈控制，但对预测神经元信号与行为之间的关系知之甚少。我们将通过寻找预测神经元信号和反应时间之间的逐次相关性来解决这个问题。我们将检查手动反应和眼球运动，以测试单个神经元的活动是否针对特定行为，或者更确切地说，提供对多种行为有用的一般预测信号。我们还将根据最近的视觉刺激试验历史，研究预测神经元信号和行为优势是否作为学习的结果而出现。我们特别假设预测神经元信号是基于预测误差（预测视觉刺激和实际视觉刺激之间的差异）更新的。此外，我们假设预测行为和预测神经元放电同样会受到近期试验历史的影响。最后，另一个“视网膜外”过程，选择性注意，已被证明可以增强感觉运动行为和调节神经元反应。一个关键的问题是，神经元调节的动态是否可以解释行为优势。我们将同时研究神经元调节的时间过程和当动物内源性地根据认知线索在两种视觉刺激之间切换注意力时所获得的表现优势的时间过程。神经元反应的注意调节的时间过程应反映神经元在内源性注意转移中所起的作用。我们的实验将为大脑如何利用感官世界的先验信息来促进感知和行动提供一个机械的视角。获得感知机制的基本信息是理解正常和异常大脑处理的必要步骤。

DESCRIPTION (provided by applicant): A great deal of evidence suggests that visual perception is affected by our predictions about the visual world. Prediction allows us to exploit and adapt to the causal, non-random structure of our visual environment, and is essential for rapid and accurate control of visually guided behavior. We previously identified neuronal signals in the monkey lateral intraparietal area (LIP) that provide a predictive representation of the direction of visual motion. The overarching issues in this new proposal are the relationship of predictive signals to behavior and the mechanisms by which predictive signals develop. Predictive motion signals in LIP seem ideally suited for feedforward control of behavior, but little is known about the relationship between predictive neuronal signals and behavior. We will address this issue by looking for trial-by-trial correlation between predictive neuronal signals and reaction time. We will examine 1 both manual reactions and eye movements to test whether the activity of single neurons is specific for particular behaviors, or rather provides a general predictive signal that is useful for multiple behaviors. We will also examine whether predictive neuronal signals and behavioral advantages arise as a result of learning based on the recent trial history of visual stimulation. We specifically hypothesize that predictive neuronal signals are updated based on prediction errors, the difference between the predicted visual stimulus and the actual visual stimulus. Moreover, we hypothesize that predictive behaviors and predictive neuronal firing will be similarly affected by recent trial history. Finally, another "extraretinal" process, selective attention, has been shown to enhance sensorimotor behaviors and to modulate neuronal responses. A key question is whether the dynamics of the neuronal modulation can account for the behavioral advantage. We will simultaneously examine the time course of neuronal modulation and the time course of the performance advantage that is gained when an animal endogenously switches attention between two visual stimuli in response to a cognitive cue. The time course of attentional modulation of a neuronal response should reflect the role that a neuron plays in endogenous shifts of attention. Our experiments will provide a mechanistic perspective on how the brain uses prior information about the sensory world to facilitate perception and action. Gaining basic information on perceptual mechanisms is an essential step for understanding normal and abnormal brain processing.