论文部分内容阅读
来自多方面的研究表明,面孔的分类和识别位于特定脑区.同时,已有行为实验研究表明,图像的空间高低频特征在面孔分类的不同范畴中起不同的贡献,例如身份更多被低频信号传递,性别被高低频共同传递,而表情更多被高频传递.然而,空间频率在面孔分类中的贡献,其表征和神经机制目前相关研究很少.利用特定癫痫患者植入颅内电极的监控期,呈现不同类型面孔图像,同时记录其颅内脑电,用事件相关电位方法考察了据认为是面孔特定成分的相关电位的潜伏期在170ms的波形(N170波形)的变化;用电极反应显著性分析考察了空间频率在不同分类特征上的贡献.结果表明,空间高频(HSF)图像的N170潜伏期显著延迟.只呈现空间低频(LSF)图像,N170的潜伏期对普通人面孔会延迟,而对熟悉的名人则没有这个差异.女性面孔诱发的N170在HSF条件下潜伏期明显晚于LSF条件,而男性面孔诱发的波形则不存在这个差异.表情在N170上没有体现出任何差异.但是基于电极的显著性分析表明,有更多的额叶电极参与了表情的加工;身份特征加工有更多电极在空间低频上表现出差异,而性别加工则空间高低频比较平衡.与以往行为结果不同的是,表情加工也有更多低频贡献,而且表情的差异可以在早达114ms的时候就发生.这符合表情信息在颞枕区域有一个快速基本加工,再传递到其他脑区的认知模型.因此,空间高低频信息在身份和性别上的贡献,可能发生在经典的面孔加工脑区,由N170表达,表情信息不由N170表达,而是在颞枕较广泛的范围内快速加工再传递到别的脑区,如额叶.这是首次利用颅内脑电就空间频率在面孔分类中的贡献的神经机制进行研究,为深入理解脑内面孔各种特征加工的动态过程提供了一个新的切入点.
Many studies have shown that the classification and recognition of faces are located in specific brain regions.At the same time, some behavioral experiments have shown that the spatial and low frequency features of images play different contributions in different categories of face classification, for example, However, there is little research on the contribution of spatial frequency to the classification of face, and its neural mechanism is very few.Using the special intracranial electrode implantation in patients with epilepsy , Different types of facial images were presented and their intracerebral electroencephalograms were recorded. Event-related potentials were used to investigate the changes of the 170 ms waveforms (N170 waveforms) of the relevant potentials that are considered to be specific components of the face. Electrode response The significance analysis showed that the contribution of space frequency to different classification features was significant. The results showed that the latency of N170 in space high frequency (HSF) images was significantly delayed. Only spatial low frequency (LSF) images were presented. The latency of N170 was delayed, While the familiar celebrities do not have this difference.The female face-induced N170 latency in HSF condition was significantly later than LSF condition, while the male face-induced The differences were not present in the waveforms.The expression did not show any difference at N170, but the significant analysis based on the electrodes showed that more frontal lobe electrodes were involved in the processing of the expression.Identity features were processed with more electrodes in the low frequency And gender processing is more balanced between high and low frequencies.With the previous behavior results are different, the facial expression processing also has more low frequency contribution, and the differences in expression can occur as early as 114ms.This is consistent with the expression information The temporal occipital region has a fast, basic process that is passed on to cognitive models in other brain regions.Therefore, the contribution of spatial high and low frequency information in identity and gender may occur in the classical face processing brain regions represented by the N170 expression information Rather than N170 expression, it is rapidly processed and delivered to other brain regions, such as the frontal lobe, over a wide range of temporal occipital lesions. This is the first study using the neural mechanism of intracranial electroencephalography (EEG) on the contribution of spatial frequency to face classification , Which provides a new entry point for the further understanding of the dynamic process of processing all kinds of features of the inner facial faces.