静息态功能磁共振成像的脑功能分区综述

发布时间：2018-03-31 08:20

本文选题：脑功能分区　切入点：静息态　出处：《中国图象图形学报》2017年10期

【摘要】：目的越来越多的研究表明,基于静息态功能磁共振成像(rs-f MRI)的大脑功能分区比传统的大脑结构分区(如AAL分区、Brodmann分区等)在功能网络构建中功能一致性更高。但现阶段对于大脑功能模块的划分较粗糙,需要更精细准确的脑功能分区,明确宏观尺度的基本功能单元。为能使脑科学领域的研究者对基于静息态功能磁共振成像的脑功能分区进行有益的探索和应用,本文对其进行系统综述。方法从rs-f MRI数据与大脑功能网络的关系出发,理清脑功能区分割的一般思路,对近几年来脑功能分区算法中出现的新思路、新方法以及对原有方法的改进做了较全面的阐述;最后总结该领域现阶段面临的问题并对未来的研究方向做了展望。结果根据脑区情况,将脑功能分区分为全脑功能分区和局部脑功能分区,并分别阐释这两方面的优势与应用。同时,将脑功能分区算法归纳为基于数据驱动和基于模型驱动两大类,并展示了各类分区算法的优势以及面临的难点和挑战。结论基于静息态功能磁共振成像的脑功能分区的研究已经取得了一些进展和有价值的研究成果,但是距离研究人脑机制,应用于脑部疾病的预防和诊断以及启示类脑科学的发展,还需要对脑功能分区方法进行更深入的研究和完善。后续研究中可将传统的分区算法和先验知识、空间领域信息、空间约束、稀疏编码、特征选择和采样学习等思想结合起来,形成融合性的脑功能分区算法,致力于更为细致准确的大脑功能分区和脑功能网络构建,解析脑的高级功能。
[Abstract]:Objective A growing number of studies have shown that, The functional partition of brain based on resting functional magnetic resonance imaging (MRI) is more consistent than that of the traditional brain structure partition (such as AAL partition, Brodmann partition, etc.), but the division of brain functional modules is rough at present. In order to enable researchers in the field of brain science to make useful exploration and application of brain functional zoning based on resting functional magnetic resonance imaging, it is necessary to define the basic functional units of the brain on a macro scale. Methods based on the relationship between rs-f MRI data and brain function network, the general idea of brain functional area segmentation is clarified, and the new ideas of brain function partition algorithm in recent years are discussed. The new method and the improvement of the original method are expounded comprehensively. Finally, the problems in this field are summarized and the future research direction is prospected. Results according to the situation of the brain region, The brain functional partition is divided into the whole brain functional partition and the local brain functional partition, and the advantages and applications of these two areas are explained. At the same time, the brain functional partition algorithm is divided into two categories: data-driven and model-based. It also shows the advantages, difficulties and challenges of all kinds of partition algorithms. Conclusion the research of brain functional partition based on resting functional magnetic resonance imaging (fMRI) has made some progress and valuable research results, but distance has been used to study the mechanism of human brain. For the prevention and diagnosis of brain diseases and the development of apocalyptic brain science, it is necessary to further study and improve the method of brain function partition. Spatial constraints, sparse coding, feature selection and sampling learning are combined to form a fusion algorithm for functional partition of the brain, which is devoted to the construction of more detailed and accurate brain functional areas and networks, and the analysis of the advanced functions of the brain.
【作者单位】：上海理工大学医学影像工程研究所;
【基金】：国家自然科学基金项目(60972122) 上海市自然科学基金项目(14ZR1427900) 上海高校青年教师培养资助计划(ZZslg15063) 微创励志创新基金项目(YS30809144)~~
【分类号】：R445.2;TP391.41

【相似文献】