Schyns Philippe G, Petro Lucy S, Smith Marie L
Centre for Cognitive Neuroimaging, University of Glasgow, 58 Hillhead Street, Glasgow G12 8QB, United Kingdom.
Curr Biol. 2007 Sep 18;17(18):1580-5. doi: 10.1016/j.cub.2007.08.048.
A key to understanding visual cognition is to determine when, how, and with what information the human brain distinguishes between visual categories. So far, the dynamics of information processing for categorization of visual stimuli has not been elucidated. By using an ecologically important categorization task (seven expressions of emotion), we demonstrate, in three human observers, that an early brain event (the N170 Event Related Potential, occurring 170 ms after stimulus onset) integrates visual information specific to each expression, according to a pattern. Specifically, starting 50 ms prior to the ERP peak, facial information tends to be integrated from the eyes downward in the face. This integration stops, and the ERP peaks, when the information diagnostic for judging a particular expression has been integrated (e.g., the eyes in fear, the corners of the nose in disgust, or the mouth in happiness). Consequently, the duration of information integration from the eyes down determines the latency of the N170 for each expression (e.g., with "fear" being faster than "disgust," itself faster than "happy"). For the first time in visual categorization, we relate the dynamics of an important brain event to the dynamics of a precise information-processing function.
理解视觉认知的关键在于确定人类大脑在何时、如何以及依据何种信息来区分视觉类别。到目前为止，视觉刺激分类的信息处理动态过程尚未得到阐明。通过使用一项具有生态学重要意义的分类任务（七种情绪表达），我们在三名人类观察者身上证明，一个早期大脑事件（N170事件相关电位，在刺激开始后170毫秒出现）会根据一种模式整合每种表情特有的视觉信息。具体而言，在ERP峰值前50毫秒开始，面部信息倾向于从眼睛向下在面部进行整合。当用于判断特定表情的诊断性信息被整合时（例如，恐惧表情中的眼睛、厌恶表情中的鼻翼、快乐表情中的嘴巴），这种整合停止，ERP达到峰值。因此，从眼睛向下的信息整合持续时间决定了每种表情的N170潜伏期（例如，“恐惧”比“厌恶”快，“厌恶”本身比“快乐”快）。在视觉分类中，我们首次将一个重要大脑事件的动态过程与精确信息处理功能的动态过程联系起来。
