基于稀疏因果分析的工作记忆障碍海马Spike/LFP功能性连接缺损机制研究

工作记忆障碍是老年性痴呆主要临床表现，本项目拟发展具有神经信息稀疏机理的稀疏因果分析方法，研究工作记忆障碍的神经信息功能连接缺损机制。 研究内容：1）建立稀疏因果模型：根据Spike的稀疏性选取稀疏空间维数，通过稀疏非负矩阵分解获取原空间记录信号的稀疏源分量，从中选取稀疏源主分量构建稀疏因果网络，计算网络特征参数。2）应用在体多通道微电极记录技术，获取在Y迷宫工作记忆和静息，正常与工作记忆障碍大鼠组海马不同模态神经信息：多通道Spikes和LFPs，获取LFP的特征频段分量。3）应用稀疏因果分析，分别计算两组大鼠海马Spikes和LFPs特征分量的稀疏因果网络参数，研究正常工作记忆海马神经信息功能连接模式，研究工作记忆障碍的信息功能性连接缺损机制。 研究意义：稀疏空间因果分析将比原空间更突出、有效表征认知信息的功能性连接，期望从信息功能连接角度在老年性痴呆症的工作记忆障碍机制研究取得突破。

Working memory dysfunction (WMD) is the clinical characteristics of Alzheimer's disease. The objective of this research is to develop a new method of sparse causal network analysis（SCNA），and to investigate the neural information disconnection mechanism of WMD via SCNA. This study includes: 1) To build a sparse causal model: the dimension of sparse space is determined from the sparseness of spikes. The sparse source components (SSCs) of the original recording signals are calculated by non-negative matrix decomposition. The sparse causal network model is built with principal components extracted from SSCs, and the feature parameters of the network are calculated. 2) Two modes of the multi-channels signals: spikes and local field potentials (LFPs) are recoded in hippocampus for normal and WMD rat groups during Y maze working memory tasks and at rest by micro-electrodes array in vivo. The characteristic frequency parts of LFPs are determined via SFFT. 3) To compute the feature parameters of the sparse causal network for two groups in the cases of working memory task and rest via SCNA. Then to investigate the functional connection of Spikes and LFPs for working memory, and the disconnection for working memory dysfunction. The causality analysis in the sparse space will represent the characterization of the functional connectivity between the cognitive signals more effectively than that in the raw recording space. A breakthrough at the aspects of working memory dysfunction mechanism of AD is expected from the view of neural information functional disconnection.