How old knowledge begets new knowledge: strategic regulation of neocortical knowledge modules and learning-induced neocortical plasticity

认知心理学和教育心理学早就发现了先验知识对新知识编码和检索的影响作用，尤其关注概念知识和图式的作用。图式直到最近才引起人们对动物和人类神经生物学文献的兴趣，而对分类的认知神经科学则了解得更多。在目前的研究计划中，我将以记忆领域图式的神经生物学最新的令人兴奋的发现为基础。重要的是，我们的研究计划旨在阐明术语图式的使用，并描述它与其他相关术语的关系——最明显的是概念，还有脚本、要点、提取的概率等等。在这些研究中，我们依赖于神经损伤患者（虚构和健忘症）和神经成像（fMRI， MEG和ERP）的数据。在神经解剖学上，我们提出，将孤立的神经模块结合成多模态集合的表征中枢的相同原理适用于概念知识，以前颞叶（ATL）为中心，对于图式，以腹内侧前额叶皮层（vmPFC）为中心。海马- vmpfc与后皮层的相互作用可介导先验图式对新信息编码的促进作用。vmPFC通过“模式实例化”来支持模式相关的记忆功能：偏向并主动维护在线上下文相关的模式，这些模式作为信息同化的支架。当患者和对照组执行图式实例化任务时，我们将使用ERP和MEG来测试vmPFC和侧后侧新皮层结构之间的皮质-皮质相互作用。我们预计健康对照的成功表现将主要取决于θ范围内的频率相干性，而这些在患者中将缺失。我们将把这些信号与从类似的概念知识任务（即哺乳动物或节肢动物等类别）中获得的信号进行比较。我们期望发现ATL而不是vmPFC在后脑新皮质活动中发挥协调作用。我们还将继续研究概念知识如何支持直接的新皮层学习。我们之前已经证明，ATL是学习新的标签-图像关联的关键，独立于海马体。其他人跟随我们的领导，证实了我们的预测，即这种学习取决于在编码新联想时相关语义的激活；这可能会导致新皮层网络的激活，而这种网络活动使其能够修改和整合新信息。利用脑磁图和我们开发的贝叶斯模型，我们计划研究新皮层活动在新皮层水平上促进概念学习的方式，以及这种学习的边界

Cognitive and educational psychology have long identified the influential role of prior knowledge on encoding and retrieval of new knowledge, with particular focus on the roles of conceptual knowledge and schemas. Schemas have only recently generated interest in both the animal and human neurobiological literature, whereas more is known about the cognitive neuroscience of categories. In the present research program I will build on recent and exciting findings on the neurobiology of schemas in the memory domain. Importantly our research program is geared towards elucidating the use of the term schema and delineating its relationship to other related terms- most notably concepts, but also scripts, gist, extracted probabilities and so forth. In these studies we rely on data from patients with neurological damage (confabulation and amnesia), and neuroimaging (fMRI, MEG and ERP’s). Neuroanatomically, we propose that the same principle of representational hubs that bind together isolated neural modules into multi-modal ensembles operate for conceptual knowledge, with the anterior temporal lobe (ATL) as the hub, and for schemas, with the ventromedial prefrontal cortex (vmPFC) as hub. Facilitation of encoding of new information by prior schemas can be mediated by hippocampal-vmPFC interaction with posterior neocortex. The vmPFC supports schema-related memory functions by “schema instantiation”: biasing and actively maintaining online contextually relevant schemas that serve as scaffold for information assimilation. Using ERP’s and MEG we will test cortico-cortical interactions between vmPFC and lateral posterior neocortical structures when patients and controls perform schema instantiation tasks. We expect that healthy controls’ successful performance would depend on frequency coherence primarily in the theta range and that these will be missing in patients. We will compare these signals to the ones derived from analogous conceptual knowledge tasks (i.e. categories such as mammals or arthropods). We expect to find that rather than vmPFC, the ATL would play a coordinating role for posterior neocortical activity. We will also continue our investigations of how conceptual knowledge may support direct neocortical learning. We have previously demonstrated that the ATL is key for learning novel label-picture associations, independently of the hippocampus. Others who followed our lead, confirming our prediction that this learning depends on activation of related semantics during encoding of the novel associations; this presumably leads to activation of neocortical networks and this network activity enables its modification and integration of novel information. Using MEG and a Bayesian model which we developed we plan to investigate the manner in which neocortical activity facilitates conceptual learning at the neocortical level, as well as the boundaries of such learning