神经系统的核心任务是在不同层次中传递与整合信息，但其传递效率又在不断改变，因此在多层次体系中研究功能的动态可变性是神经科学的关键问题之一。适应与训练是功能可变性中两种截然相反的表现形式，近年来视觉适应的研究揭示出不同层次间非均一的、选择性级联放大的复杂形式，我们同时发现方位适应与训练具有双向性，但训练的层次性表现形式尚不清晰。本申请书提出，在猫初级视觉通路，采用新型的倾向性概率的方位刺激范式，研究从视网膜、外膝体、17区到21a区的方位训练的生理学机制，并带动新型碳纳米管电极的设计与深入应用，充分发挥其柔性、微型、生物兼容性好、性质稳定的特点，在以上多个脑区同步、长期慢性单细胞记录并比较方位训练与适应机制，着重其起源和层次性传递方式，并利用自组织模型进行模拟及动物实验验证，从而揭示方位训练的层次性传递机制，为基于类脑模拟的神经网络提供生理学基础，并推动国产新一代电极检测技术的研发与应用。

The major goal of the nervous system is to transfer and integrate information between different levels. However, the integration mechanism is always modified by continuous stimulations, thus it is a central theme to investigate the dynamic plasticity in a multi-level system. Adaptation and training are two opposite functional plasticity. Recent studies demonstrated that visual adaptation was non-homogeneous and selectively cascading between different levels, and our recent studies found the dual mode of orientation adaptation and training, but the process of the visual training still remains unclear. We propose to investigate the orientation training from retina, LGN, area 17 to area 21a in cat primary visual pathway, using the new biased-orientation protocol, by the new carbon nanotube electrode. This electrode is flexible, tiny, bio-compatible and stable, thus will enable the simultaneous chronic single-unit recordings of multiple visual areas. We will focus on the relationship between orientation training and adaptation, its origin and cascading effect between different levels. At the same time, self-organization model will be applied to simulate this process, and then be confirmed by animal experiments. These studies will help to understand the mechanism of the hierarchical information processing, thus provide direct physiological basis of brain inspired artificial neural networks. Furthermore, it will enable the establishment and application of the next generation nano-sensing technology.