Assessment of parallel movement plans by probing spatial attention

我们的环境不断为我们提供多种机会和行动要求。在任何给定的时刻，我们必须选择所有可能的动作之一，并指定相应的移动度量。有证据表明，大脑同时并行地准备多个动作，并且动作选择是由嵌入在自下而上驱动的感觉运动处理中的持续竞争引起的，该处理受到自上而下的决策相关信息的偏见。人类参与者的到达轨迹通常反映了到几个当前相关目标的直接轨迹之间的平均值;运动行为的这一方面被解释为指示执行的运动是由个体运动计划的平均值产生的。因此，这种解释假设不同行动的所有方面都是并行表示的。然而，平均行为有时似乎取决于战略考虑。例如，当目标相距很远时，或者当运动必须非常快时，放弃轨迹平均;这种策略行为变化提高了参与者的整体表现。使用这种策略已经被用来指示默认情况下移动轨迹不是平均的。相反，只计算一个详细的移动计划，当这是战略上有利的时候，这个单一的计划可以作为一个平均值得出。因此，这种理论立场假设最终运动目标的平行表示，而不是运动计划的平行表示。解决这一争论的核心挑战是找到令人信服的衡量标准，表明几个运动计划（而不仅仅是目标）当前是否有效。我们在这里提出了一系列的实验来应对这一挑战。为此，我们采用运动目标和注意力部署之间的良好关系。注意力转移到一个或几个连续的运动目标已经在运动开始之前，在目标相比，不相关的位置表示在增强的感知辨别性能。我们扩展这种实验方法来探测多个运动目标的手达到的并行表示。关键的是，我们探测注意力部署到边界定义的位置，如障碍物附近的区域，沿着潜在的相关运动轨迹。作为第二种方法，我们通过运动适应范式诱导多个潜在相关的轨迹。最后，我们解决自上而下的运动计划选择方面，以阐明这些方面如何影响自下而上的感觉信息的平均。总之，拟议中的实验将提供实质性的和令人信服的证据，运动规划水平的基础上并行表示。

Our environment constantly presents us with multiple opportunities and demands for action. At any given moment, we must select one of all possible actions, and specify the corresponding movement metrics. Evidence suggests that the brain simultaneously prepares multiple actions in parallel, and action selection results from continuous competition embedded in bottom-up driven sensorimotor processing that is biased by top-down decision-relevant information.However, there is ongoing debate about which aspects of actions are actually represented in parallel. Reaching trajectories of human participants usually reflect an average between the direct trajectories to several currently relevant targets; this aspect of motor behavior has been interpreted as indicating that the executed movement results from averaging of individual movement plans. This interpretation, thus, assumes that all aspects of the different actions are represented in parallel. However, averaging behavior appears to sometimes depend on strategic considerations. For instance, trajectory averaging is abandoned when targets are far apart, or when movements have to executed be very fast; this strategic behavior change improves participants’ overall performance. The use of such strategies has been taken to indicate that movement trajectories are not, by default, averaged. Instead, only a single detailed movement plan would be computed, and this single plan could be derived as an average when this is strategically advantageous. Thus, this theoretical stance assumes parallel representation of final movement goals, but not of movement plans that specify aspects such as the movement’s trajectory.The central challenge to solving this debate is to find convincing measures that indicate whether several movement plans (as opposed to just goals) are currently active. We propose here a series of experiments that address this challenge. To this end, we employ the well-established relationship between movement goals and attentional deployment. Attention is shifted towards one or several sequential motor goals already prior to movement initiation, expressed in enhanced perceptual discrimination performance at target as compared to irrelevant locations. We extend this experimental approach to probing the parallel representation of multiple motor goals for hand reaches. Critically, we probe the attentional deployment to trajectory-defining locations, such as regions near obstacles, along potentially relevant movement trajectories. As a second approach, we induce multiple potentially relevant trajectories through a motor adaptation paradigm. Finally, we address top-down aspects to motor plan selection to elucidate how such aspects affect the averaging of bottom-up sensory information. Together, the proposed experiments will provide substantive and cogent evidence about which levels of movement planning underlie parallel representation.