融合时-空脑信息的理论、方法及其在人类工作记忆及老化研究中的应用

项目名称： 融合时-空脑信息的理论、方法及其在人类工作记忆及老化研究中的应用

项目编号： No.91232725

项目类型： 重大研究计划

立项/批准年度： 2013

项目学科： 影像医学与生物医学工程

项目作者： 尧德中

作者单位： 电子科技大学

项目金额： 120万元

中文摘要： 工作记忆及其老化研究的深入需要有效整合多模态信号提供的时-空互补信息，但目前还缺乏能对功能-结构、功能-功能信息进行有效整合的理论与方法。本项目将基于信息科学理论与方法，对多模态时-空脑信息融合的理论、方法及结合记忆研究的应用进行系统性的研究，发展和建立满足不同情况需求的,系列无创多模脑信息融合的方法与技术，为记忆及其老化机制的深入研究提供新的技术手段。在理论方法方面，将以维恩关系图为纽带来对多模态数据间的关系进行统一的描述，发展以信息理论为主线的，对多模信息的共性和特异性信息同时关注的系列融合方法；在应用方面，将获取具有较长时间跨度的不同年龄人群的结构、功能及行为信息，然后利用新发展的多模时-空信息融合方法为工作记忆及其老化研究提供有较高时空区分度的动态信息，以对记忆相关机制进行从单模态到多模态，从点（功能定位）到面（网络、有效连接）的系统研究，揭示工作记忆及其老化过程的深层神经机制。

中文关键词： 多模态；信息融合；信息理论；记忆老化；

英文摘要： As for the study of working memory and its aging process, it is necessary to deeply mine the complementary information of multi-modality physiological signals. However, the analysis techniques that can effectively integrate the information of multi-modality signals have not been explored yet. In this project, to satisfy the urgent requirement of working memory and its aging, we will build a series of theories and methods for the fusion of non-invasive multi-modality brain signals, whose core is the basis of information theory. As for the aspect of theory and methology, we will use the Venn graph to unify the relationships existed among multi-modality signals, which may provide a very consistent framework to perform the further fusion. Based on information theory, we will develop a series of fusion methods to mine the complementary information of multi-modality signals revealed by the Venn graph, among which not only the common information but also the specified information of certain modality can be extracted as desired. After the fusion methods have been developed, we will apply them to the neural mechanism study of working memory and its aging process. Compared to previous similar working memory studies, two novel points will be provided in this project: 1) The population has a very wide range cover, which wil

英文关键词： Multi-modal；information fusion；information theory；memory aging；