Perturbing gaze shifts reveals new visuo-motor encoding mechanisms

人类擅长做出快速、准确的动作，这涉及到同时运动的功能耦合的身体部分。大脑如何协调如此复杂的运动一直是我的实验室的一个主要关注点，作为一个实验模型，我们研究了眼睛+头部的协调运动，这种运动通常会产生视轴的快速定向运动，称为凝视眼跳。*中脑的上丘(SC)是凝视控制的关键结构。经典的假设是，SC包含一个运动地图，它根据一种编码方案驱动快速(扫视)凝视转移，其中地图上的一个位置编码一个凝视转移矢量。我们称之为“一站式一扫视”假说。我们有初步证据表明，当一个人使用一系列由注视周期分隔的快速眼跳来观察环境时，这一假设是不成立的。在这种情况下，SC对整体位移而不是单个扫视进行编码。这一与SC函数主导假设的关键突破需要得到验证，这是我们在这里建议做的。如果我们确认失败，那么这将需要对大脑如何控制我们对环境的看法的现有模型进行关键的重新评估。*观察环境的目的显然是为了“看”。这个简单的日复一日的术语隐藏了一个极其复杂的大脑功能集合，这些功能是感知的基础。首先，我们需要将视觉与眼睛的运动联系起来。这种联系是必要的，因为：1)当我们进行3-5次/秒的扫视时，尽管我们感觉到的物体在视网膜上的位置发生了变化，但我们的视觉世界在扫视过程中似乎是稳定的；以及2)在每次扫视后，空间上固定的注意力轨迹在视网膜上占据不同的位置，但在内部必须被感知为空间上不动。*有趣的是，大脑的视觉系统是以视觉地图的方式组织起来的，这种方式类似于SC的运动地图。因此，视觉地图上的一个地点编码了视觉空间中的一个小位置。这就是所谓的“一个地点，一个接受野”假说。与凝视运动系统类似，有证据表明这一假设可能会失败。事实上，视觉地图中的神经元可以对视觉世界中任何地方呈现的视觉刺激做出反应，这取决于即将到来的扫视的方向和幅度。因此，视觉地图上的一个站点可以编码无限数量的空间位置！这种与经典思维的彻底决裂需要被理解为理解远见。我们计划在这里通过研究当预期的眼跳受到干扰时，视觉神经元如何对视觉刺激做出反应来做到这一点。综合起来，运动和视觉系统中大脑组织的关键原则可以被称为“一处一功能”假说。我们有这些系统如何工作的模型，但当它们不起作用时，如果我们要了解大脑的功能，以及当我们扫描环境时如何看到，理解为什么是至关重要的。**

Humans are skilled at making rapid, accurate movements that involve functionally coupled body segments that move simultaneously. How the brain orchestrates such complex movements has been a major focus of my lab and, as an experimental model, we have studied coordinated eye + head movements that routinely generate rapid orienting movements of the visual axis called gaze saccades. ***The midbrain's superior colliculus (SC) is a critical structure in gaze control. The classic hypothesis is that the SC contains a motor map which drives rapid (saccadic) gaze shifts according to a coding scheme wherein one-site on the map encodes one gaze shift vector. We call this the “one-site, one- saccade” hypothesis. We have preliminary evidence that this hypothesis fails when one looks about the environment using a sequence of rapid gaze saccades separated by periods of fixation. In this context the SC encodes the overall displacement not individual saccades. This critical break with the dominant hypothesis of SC function needs to be verified which we propose to do here. If we confirm failure then this will call for a critical re-evaluation of current models of how the brain controls how we look about the environment. ***The objective of looking about the environment is obviously to ‘see'. This simple day to day term hides an extraordinarily complex ensemble of brain functions that underlie perception. For one thing we need to link vision with movements of the eyes. Such a link is necessary because: 1) our visual world appears stable across saccadic gaze shifts despite a perceived object changing position on the retina as we make saccades 3-5 times/sec; and 2) a spatially-fixed locus of attention occupies a different position on the retina after each saccade but must internally be perceived as spatially immobile. *** ***Interestingly, the brain's visual system is organized in visual maps in a manner analogous to the SC's motor map. Thus, one site on a visual map encodes one small location in visual space. This is called the “one-site, one-receptive-field” hypothesis. Similar to the gaze motor system there is evidence that this hypothesis can fail. Indeed, a neuron in a visual map can respond to a visual stimulus presented anywhere in the visual world depending on the direction and amplitude of the upcoming saccade. Thus one site on a visual map can encode an infinite number of spatial positions! This radical break with classic thinking needs to be understood to understand vision. We plan to do this here by studying how visual neurons respond to visual stimuli when the intended gaze saccade is perturbed. ***Together, the key principles of brain organization in the motor and visual system can be referred to as the “one-site, one-function” hypotheses. We have models of how these systems work but when they don't work, it is critical to understand why if we are to understand brain function and how we see when we scan our environment.**