Control of memory-guided attention

在最近的研究中，我们发现，相对于随机字母搜索阵列，视觉搜索在重复中速度更快，这种行为反应时间的情境促进伴随着并与增强的侧侧ERP反应相关，这些反应在刺激开始后80 - 180毫秒达到峰值（N1pc成分，随后是侧化的N2pc和CDA，分别从200和400毫秒开始）。这些发现证实了现有的统计学习理论，根据这些理论，重复出现的搜索模式会导致更有效地将注意力分配给潜在的目标项目（由N2pc反映）。然而，我们的新发现也表明了这些解释的延伸，例如统计学习经验增加了潜在目标项目（N1pc）的注意优先级，关键的是，甚至在自愿注意分配（N2pc）之前，同时也对搜索引导工作记忆（CDA）中表示的（目标/干扰物）信息的数量产生影响。在提议的项目中，我们将使用视觉搜索并研究N1pc和CDA的感觉认知机制（重复排列更有效），使用行为反应时间结合时间精确的脑电图，包括脑电图信息的功能MRI。我们怀疑，暂时早期的N2pc标志着在有助于长期记忆目标-环境关系和/或视觉引导的空间注意的电路中编码的空间地图的经验依赖变化，并实例化了自愿目标选择（N2pc）和最终注意抑制（PD）的偏见。后者发生在目标位置改变后（从一个半球到另一个半球），如果最初的注意定向过程不能检测到与任务相关的目标项目。关于CDA：我们将测试一个新颖的想法，即从学习搜索数组中获得的注意定向与使用可见箭头提示的注意定向一样有效，从而在横向视觉搜索中引发CDA。此外，我们还询问CDA是否反映了强制性的工作记忆编码，以及学习搜索数组中CDA的增强是否标志着分心物和/或目标项目的工作记忆增强。

In recent work we found that visual search is faster in repeated relative to random letter search arrays, with this contextual facilitation of behavioral reaction times being accompanied by, and correlating with, enhanced lateralized ERP responses peaking already 80 - 180 milliseconds post-stimulus onset (N1pc component and followed by the lateralized N2pc and CDA, starting at around 200 and 400 ms, respectively) over parieto-occipital electrode sites. These findings confirm existing statistical-learning accounts according to which re-occurring search patterns lead to more effective attentional allocation towards potential target items (reflected by the N2pc). However, our new findings also suggest an extension of these accounts such that statistical-learning experiences increase the attentional priory of potential target items (N1pc), critically, even before voluntary attentional allocation (N2pc), while also exerting an influence on how much (target/ distractor) information is represented in search-guiding working memory (CDA). In the proposed project, we will use visual search and investigate the sensory-cognitive mechanisms underlying (the in repeated arrays more effective) N1pc and CDA using behavioral reaction times in combination with temporally-precise EEG, including also EEG-informed functional MRI. We suspect that the temporally-early N1pc marks experience-dependent changes of spatial maps encoded in circuits contributing to long-term memory of target-context relations and/ or visually guided spatial attention and instantiating a bias for voluntary target selection (N2pc) and eventual attentional suppression (PD), the latter after a target location change (from one hemifield to the other) in a learned contextual array and if the initial attention-orienting process was not capable to detect the task-relevant target item. Concerning the CDA: we will test the novel idea that attentional orienting from learned search arrays is as effective as that with visible arrow cues and thus elicit CDAs during lateralized visual search. Moreover, we ask whether CDAs reflect obligatory working-memory encoding and whether the enhanced CDA in learned search arrays marks enhanced working memory for distractor and/ or target items.