人类面孔和躯体信息整合的神经机制

Previous neuroimaging studies have demonstrated the existence of brain areas that selectively respond to visually presented faces and bodies, which are thought to subserve the processing of faces and bodies of people respectively. However, in real world, we see “whole” people and not just isolated faces and bodies, and the question remains of how information from these two categories of stimuli is integrated in the brain. Are people represented by distributed face- and body-selective neurons, or are there neural populations that actually code for whole individuals? Based on a recent monkey study as well as observations from the applicant’s own, the current project proposed a neuroanatomical model for face-body integration: the basic and the dynamic shape of people is gradually evolved from the face-selective areas in the ventral and dorsal visual pathway respectively; the integration of identity, emotion and eye-gaze direction information from faces and bodies differentially relies on fusiform-ventral anterior temporal, posterior superior temporal sulcus-amygdala, and intraparietal sulcus-posterior superior temporal sulcus face-selective areas. To test this model, the current study will thoroughly localize face- and body-selective areas in ventral and dorsal pathway using a dynamic localizer, and critically test the selectivity and specificity of these areas to the information conveyed by people using the representational similarity analysis. Whether the model be validated or not, the study will provide information for human face and body integration at neural level, and will have general implications for our understanding of how our brain constructs meaningful objects from different but relevant visual inputs.

成像研究表明大脑中存在对面孔和躯体有选择性反应的脑区，它们被认为分别加工人的面孔和躯体。然而，现实中我们看到的是整个人而不是单独的面孔和躯体，大脑如何整合来自于面孔和躯体的信息？人是由面孔和躯体选择神经元分布表征的，还是有专门编码人的神经元群落？基于以往猴子以及申请人自己的研究，本项目提出一个面孔和躯体整合的神经解剖模型：人的基本形状和动态形状分别从腹侧通路和背侧通路的面孔加工区演变而来；来自于面孔和躯体的身份，情绪，注视方向信息的整合分别依赖梭状回－腹侧颞叶前部，后颞上沟－杏仁核，顶内沟－后颞上沟面孔区。模型验证：项目采用动态定位技术全面定位大脑腹背侧上的面孔和躯体选择区，采用表征相似性分析方法分析这些区域对人携带的信息的选择性和特异性。无论模型是否被验证，项目都能为人类面孔和躯体信息整合的神经机制提供信息。研究对于我们理解大脑如何从不同但相关的视觉输入中构造有意义的物体也具有普遍意义。

采用单独的面孔刺激和躯体刺激，先前研究发现人类大脑中存在相对独立的面孔加工系统和躯体加工系统。然而，现实生活中面孔和躯体总是一起出现，大脑如何整合面孔加工系统和躯体加工系统的信息来表征人？本研究围绕这个问题系统地开展了三项研究。研究一通过记录和分析个体在高空间分辨率磁共振扫描仪内观看“人”时的 大脑活动确定大脑用哪些脑区表征人。研究发现“人”加工区分布在梭状回中部，枕下回、枕外沟以及颞上沟后部面孔选择区和躯体选择区的重叠处。研究二通过考察“人”加工区对包含不同面孔和躯体信息的六类刺激的反应模式确定它们对“人”加工的具体贡献。研究发现：枕下回主要负责加工面孔与躯体的空间构造关系；枕外沟主要负责加工躯体信息；梭状回中部主要负责加工人的面孔信息；颞上沟后部主要负责加工刺激的生物性信息。研究三在前两项研究的基础上，进一步采用高时间分辨率脑磁图技术确定大脑加工“人”的时间进程。研究结果表明“人”的加工包括三个阶段: 首先是颞上沟后部对刺激的生物性进行判断（刺激呈现后20-30ms），然后是枕下回对人模板符合性进行判断（70-80ms），最后是梭状回（80-90ms）和枕外沟（100-110ms）分别对人的面孔和躯体信息进行细致加工。本研究首次采用高空间分辨率功能磁共振和高时间分辨率脑磁图系统探索了大脑如何表征由面孔和躯体组成的“人”。研究结果为人类大脑如何整合面孔和躯体信息表征“人”提供了理论框架。此外，本项目执行过程中催生了一套评估面孔和躯体整合能力的多模态脑成像方法，该方法可以用于精神疾病、神经退行性疾病患者社会认知能力的评估。