Neural correlates of visual ambiguity and meta-perceptual confidence rating

我们感官所能获得的信息是不完整和模棱两可的。我们的知觉系统通常将这些自下而上的感觉信息与自上而下的知觉记忆结合起来，以解决这种模糊性，并构建稳定可靠的知觉。在观察一个模棱两可的人物时，我们的知觉变得不稳定，在两种同样可能但相互排斥的解释之间交替。我们最近发现了两个事件相关电位成分(P200和P400)，它们分别由模糊图形和消解图形变量(事件相关电位模糊效应)引起的小波幅和大波幅。这些异常强烈的效果在非常不同的视觉类别和复杂程度(运动、语义、情绪面部表情和字母/数字)中得到概括。最近对Necker立方体刺激的一项研究显示，关于知觉加工的预测编码模型得到了完美的证实。该模型表明，P200与自上而下对知觉过程的贡献相关。P400似乎反映了自下而上和自上而下因素的整合，从而反映了知觉结果的可靠性。这些发现和介绍的模型是这个由三个项目组成的拨款提案的起点：项目A：测试跨视觉类别和复杂性水平的概括性到目前为止，ERP歧义效应和预测编码模型之间的关系只在Necker立方体刺激下显示。我们计划测试这些符合模型的发现是否适用于上述其他刺激类别和复杂程度。项目B：ERP歧义影响测试SSD患者的变化精神分裂症谱系障碍(SSD)患者在整合自下而上的感觉和自上而下的信息方面存在缺陷。我们计划测试这种整合缺陷是否反映在ERP模糊效应的偏离模式中。项目C：关于神经源的事件相关电位模糊效应的概括性测试到目前为止，事件相关电位模糊效应的空间位置是基于脑电电压图。这个子项目的目标是，使用同步的EEG/fMRI记录：(1)明确不同刺激类别和复杂程度的相似ERP头皮分布是否基于模型预测的相同潜在神经来源。(2)研究SSD患者和对照组之间知觉歧义消解过程的脑区及其差异。本课题在神经生物学、知觉心理学和精神病学之间的研究结果可能会促进对事件相关电位模糊效应的神经和功能特征以及知觉的预测性编码方法的理解。鉴于ERP多义性效应的规模很大，本项目的结果可能在未来的临床应用中具有良好的前景。

The information available to our senses is incomplete and ambiguous. Our perceptual system normally integrates this the bottom-up sensory information with top-down perceptual memories in order to resolve this ambiguity and to construct stable and reliable percepts. During observation of an ambiguous figure our perception becomes unstable and alternates between two equally probable but mutual exclusive interpretations. We recently found two ERP components (P200 and P400) with small vs. large amplitudes evoked by ambiguous figures vs. disambiguated figure variants (ERP Ambiguity Effects). These unusually strong effects generalize across very different visual categories and complexity levels (motion, semantics, emotional face expressions and letters/numbers). A recent study with Necker cube stimuli revealed perfect confirmation of a predictive coding model about perceptual processing. The model indicates that the P200 correlates with top-down contributions to the perceptual process. The P400 seems to reflect the integration of bottom-up and top-down elements and thus the reliability of the perceptual outcome. These findings and the introduced model are the starting point of this grant proposal consisting of three projects:Project A: Testing for generalizability across visual categories and complexity levelsSo far, the relation between the ERP Ambiguity Effects and the predictive coding model has only been shown for Necker cube stimuli. We plan to test whether these model-conform findings generalize across other stimulus categories and complexity levels mentioned above.Project B: Test for ERP Ambiguity Effects alterations in SSD PatientsPatients with schizophrenia spectrum disorder (SSD) show deficits in integrating bottom-up sensory with top-down information. We plan to test whether this integration deficit is reflected in a deviated pattern of ERP Ambiguity Effects. Project C: Testing for generalizability of the ERP Ambiguity Effects concerning neural sourcesThe spatial locations of the ERP Ambiguity Effects are so far based on EEG voltage maps. The aims of this subproject are, using concurrent EEG/fMRI recordings: (1) To specify whether similar ERP scalp distributions across different stimulus categories and complexity levels are based on the same underlying neural sources, as predicted by the model. (2) To identify brain areas underlying perceptual disambiguation processes and respective differences between SSD patients and controls. The results of this project at the interface between neurobiology, psychology of perception and psychiatry may advance the understanding of the neural and functional features of the ERP Ambiguity Effects and predictive coding approaches to perception. Given the large sizes of the ERP Ambiguity Effects, the results from the current project may have a good prospect for future clinical applications.