Depth perception is an important aspect of normal vision. The three-dimensional world collapses into two-dimensional images at the retina. It is well accepted that depth information is extracted in the viusal central nervous system from the slight difference between images from left and right eyes. Much is known about how individual neurons encode the binocular disparity information with single electrode single-unit recording techniques. However, how networks of neurons represent binocular disparity information in population activity has rarely been investigated experimentally. We will perform high resolution functional brain imaging to examine the distribution of binocular disparity-coding regions at different cortical processing stages, their functional connectivity and relationship to other functionally defined regions. With improved multi-electrode techniques and brain window, we plan to implant imaging-guided electrode array to simultaneously monitor neuronal activity at multiple cortical areas (e.g., V1 and V2) of behaving non-human primates. These experiments will allow us for the first time to examine the flow of information for binocular depth perception between different cortical areas and the coding of binocular disparity information at multiple stages of visual processing. This work will help to establish how the perception emerges from the activity of neural circuits, and will help the development of treatment for illnesses that impair normal visual sensation.
深度视觉是正常视觉的重要组成部分。三维世界在视网膜上被压缩为两幅不具有深度信息的二维图像。在视觉中枢中,神经元比对左眼和右眼图像之间的细微差异来重建深度信息。大量使用的单电极电生理记录手段能详尽研究单个神经元对双眼视差刺激的反应, 但无法获知群体神经元是如何作为一个整体来处理深度视觉信息的。首先我们将运用脑功能成像来定位不同视区中双眼视差功能区的确切位置并获得这些功能区之间的神经网络连接特点。然后在深入研究这些深度视功能结构和其他功能结构的关联后,再依据脑成像结果在大脑窗口中有针对性的植入高密度电极阵列。通过同步记录清醒非人灵长类的多个脑区(例如初级视区和次级视区)中的神经元群体活动,我们不仅能首次揭示深度视觉信息在不同视区之间的传递规律,而且会获知不同脑区中双眼视差信息编码策略的异同。研究结果将显著提高我们对神经元的电活动和视知觉之间关联的理解,并将为视知觉损伤的治疗提供新思路
本研究发现次级视觉脑区和群体神经活动在深度视觉感觉的形成中起着关键的作用。在麻醉和清醒非人灵长类实验动物进行的大脑成像结果表明双眼匹配特征抽提这一重要深度视觉信息处理步骤发生在次级视区中能特异性处理远深度感觉或近深度感觉的双眼水平视差大脑神经功能构筑。这种存在于次级视觉脑区的功能结构能通过整合相关的双眼水平视差群体神经活动实现有效的抑制错误双眼匹配信息从而产生深度视觉感觉。
{{i.achievement_title}}
数据更新时间:2023-05-31
基于SSVEP 直接脑控机器人方向和速度研究
基于公众情感倾向的主题公园评价研究——以哈尔滨市伏尔加庄园为例
多源数据驱动CNN-GRU模型的公交客流量分类预测
采用深度学习的铣刀磨损状态预测模型
涡轮叶片厚壁带肋通道流动与传热性能的预测和优化
视觉信息处理中的神经元群体编码理论研究
面向视觉质量优化的深度学习视频编码
视觉信息的群体动态时空编码和选择性注意机制
Gamma振荡对视觉信息群体编码的动态影响