Dynamic task dependent decoding of visual motion information during continuous behavior

为了在日常生活中采取行动和做出反应，我们必须不断整合来自感官系统的信息，以解决广泛的不断变化的任务。尽管对神经信息整合的路径有着巨大的见解，但人们对如何实现解码视觉信息以解决自然场景中复杂任务所需的灵活性知之甚少。在实验室里，这种专注的特性通常是孤立地研究的，并做出离散的决定。我的目标是通过一系列创新的实验和分析来克服这些局限性。我将记录数据在连续和自然的行为和刺激在三个互补规模的神经处理：同时单细胞反应的区域MT和MST的猕猴大脑，大规模的皮层电路（人类功能磁共振成像）和由此产生的行为。我的目标是将一个通用的统计模型应用于这些数据，这将为神经过程提供严格的见解。该项目将为未来的研究奠定基础，在未来的研究中，神经科学数据可以在复杂的自然环境中收集和分析。这将允许理解连续自然行为中的神经计算，并减少当前对受试者，特别是研究中的动物的限制。

In order to act and react in our daily lives, we have to continuously integrate information from our sensory systems to solve a broad range of constantly changing tasks. Despite tremendous insights into the paths in which neural information is integrated, little is known how the required flexibility in the decoding of the visual information to solve complex tasks in natural scenarios is achieved. In the laboratory, such attentive properties are typically investigated in isolation with discrete decisions being made. I aim to overcome these limitations using a set of innovative experiments and analyses. I will record data during continuous and naturalistic behavior and stimulation at three complementary scales of neural processing: simultaneous single cell responses within areas MT and MST of the macaque brain, large scale cortical circuits (human fMRI) and the resultant behavior. I aim to apply a common statistical model to this data that will provide rigorous insights into the neural processes. The project will lay the groundwork for future research in which neuroscientific data can be collected and analyzed in complex, natural environments. This will allow understanding neural computations in continuous natural behavior and reduce current constraints imposed on the subjects and especially animals under research.