视觉物体的整体形状是由局部细节特征构成，大脑对物体局部和整体方位信息的加工与整合，是视觉科学的核心问题之一。灵长类可以毫不费力地在识别物体整体轮廓方位的同时，看清其局部的方位特征，这背后复杂的神经环路机制仍不清楚。申请人前期发表的研究工作首次发现，在灵长类V4脑区就同时存在两类不同空间频率偏好的神经元，分别编码局部和整体方位信息。但是，这两类神经元在神经环路以及编码机制上，是如何分别实现局部方位特征和整体轮廓方位的整合的？二者是如何交互关联的? 这是本申请的核心科学问题。本项目利用非人灵长类猕猴为动物模型，结合双光子成像与内源性光学成像方法，在树突、单细胞、群体细胞等多尺度上，深入系统地研究V4这两类神经元在局部和整体方向信息加工过程中的神经编码和环路机制，研究结果将帮助我们认识和理解外部物体在灵长类视觉大脑中的重构原理，也为神经计算，类脑人工网络模型和智能信息处理提供重要的理论依据。

Local and global orientation integration is a fundamental question in visual science. Most previous psychological and neuronal studies focused on the mechanism of collinear orientation integration. But human and non-human primates can effortlessly recognize the global shape of an object, while at the same time visually resolving the local orientation features, especially when the local and global orientations are not collinear. In our recent work, we find higher-acuity neural clusters represent local features, accompany with lower-acuity neural clusters represent global features in V4 of non-human primates. This finding suggests that we do not necessarily need to employ complex recurrent loops across the visual hierarchy to connect global features in the higher visual areas and local information in the earlier visual cortices, for seeing both global and local orientation features clearly. In this study, using intrinsic optical imaging at population level and two-photon imaging at single cell and dendritic spine resolution, we wish to further investigate the exact neural/circuit mechanisms underlying local and global orientation processing by the two types of neurons in visual cortex V4 of rhesus macaque. The findings will help us to critically further our understanding on the cortical integration on how our visual brain works and what necessary neural/circuit mechanisms underlie human perception.