CASIA OpenIR  > 脑图谱与类脑智能实验室  > 脑网络组研究
猕猴脑网络组图谱绘制及其应用
陆玉恒
2024-05-12
Pages126
Subtype博士
Abstract

猕猴作为实验动物,因其与人类在基因、大脑组织和行为等方面高度相似,被广泛用于脑科学研究。在猕猴大脑结构与功能的探索中,脑图谱发挥着重要作用。目前,大多数猕猴脑图谱基于传统的细胞构筑方式制作,其绘制标准存在一定主观性,而且在绘制过程中仅利用了脑区的局部属性。然而,大脑是由大量跨越脑区的连接纤维组成,这些连接模式反映了大脑功能。因此,基于连接绘制的猕猴全脑图谱对于理解猕猴大脑至关重要。鉴于目前仍缺乏猕猴脑连接图谱这一限制,本文以猕猴脑图谱绘制为核心,利用弥散磁共振数据构建了基于连接的猕猴脑网络组分区图谱。为验证分区图谱,我们运用组织学和示踪数据,并将多模态数据集成到图谱空间,形成了一个包含多模态信息的猕猴脑图谱。在应用方面,我们使用猕猴脑网络组图谱研究了物种间大脑连接差异和图谱个体化方法。本文的主要内容及创新点归纳如下:
(1) 基于弥散磁共振图像的猕猴脑网络组图谱分区绘制。大脑是由连接构成的整体,连接模式与脑区的功能密不可分。因此,本文基于高空间分辨率和高角度分辨率的猕猴离体弥散磁共振数据,提取全脑结构连接特征,建立了猕猴大脑的自动化分区流程。在此基础上,我们绘制了具有群体水平和精细划分的猕猴脑网络组图谱。该图谱将猕猴大脑划分为皮层上 256 个和皮层下 58 个亚区。另外,通过对一组公开数据集的重复性验证,我们验证了该绘制流程和结果的鲁棒性和可重复性。此外,与其他已有的猕猴脑图谱相比,猕猴脑网络组图谱分区能够更准确表达大脑连接的拓扑关系。
(2) 猕猴脑网络组图谱的多模态信息集成与验证。已有的猕猴脑图谱往往仅包含一种模态的信息,无法全面反映大脑的整体情况。因此,在完成猕猴脑网络组图谱分区的基础上,本文结合多模态磁共振数据,建立了每个脑区间的结构连接和功能连接。随后,我们将来自同一只猕猴的染色切片数据和示踪注射数据整合到脑网络组图谱空间,系统地验证了猕猴脑网络组图谱的可靠性。该图谱集成了结构连接、示踪连接和组织学等多模态信息,弥补了过去猕猴图谱仅反映大脑单一模态属性的不足,为全面、多角度认识猕猴大脑提供了重要途径。
(3) 基于人类和猕猴脑网络组图谱的脑连接比较研究。在进化过程中,不同物种的大脑连接模式发生了改变。通过对比人类和猕猴大脑连接模式的差异,我们能够更好地理解人类独有的高级功能。基于猕猴和人类脑网络组图谱,本文首先分析了猕猴和人类左右大脑之间的连接差异模式。接着,我们以猕猴和人类脑网络组图谱作为比较单元,描述了人类与猕猴在大脑连接上的物种间差异程度。最后,通过对连接差异与人脑基因表达数据的联合分析,我们筛选出了最有可能在进化过程中驱动这些连接差异的基因集,并通过基因富集分析发现该基因集与人类的智力和一些精神疾病有密切关系。这可能表明,人类在进化过程中展现出的与猕猴大脑的连接差异同时导致了人类智力的提升和部分精神疾病的出现。

(4) 猕猴脑图谱的个体化方法研究。在猕猴脑图谱的实际应用中,传统方法通常采用基于形态学的图像配准技术,将脑图谱直接映射到个体脑上,却忽视了猕猴大脑间同样存在与人类相似的个体差异。为了解决这一问题,本文利用组水平的猕猴脑网络组图谱作为先验信息,提出了一种基于大脑连接的猕猴图谱个体化方法。该方法将每个个体的特征信息与组平均的特征信息共同映射到一个低维隐空间,随后采用贝叶斯估计的方式对猕猴大脑进行个体分区。与其他个体化方法的比较结果显示,本文提出的方法能更有效地反映个体连接信息的分布情况。

Other Abstract

As an experimental animal, macaque is extensively employed in neuroscience re-search due to its high similarity to humans in terms of genetics, brain structure, and behavior. Brain atlases play a crucial role in exploring the structure and function of the brain. Currently, most macaque brain atlases are constructed based on traditional cytoarchitectonic methods, which inherently entail subjectivity in standardization and utilize local properties of brain regions during the mapping process. However, the brain comprises extensive connections across the whole brain, and connectivity constrains the function of the brain. Therefore, the construction of macaque whole-brain atlases based on connectivity is vital for understanding the macaque brain. This paper focuses on macaque brain atlas construction and utilizes diffusion magnetic resonance imaging data to establish a connectivity-based parcellation of the macaque brain. To validate the parcellation atlas, histological and tracing data are employed, and multimodal data are integrated based on the atlas, making a macaque brain atlas with multimodal informa-tion. In terms of applications, the macaque brain atlas delineated in this paper is utilized to investigate the brain connectivity divergence between macaques and humans, and to work as a prior in the proposed individualized parcellation method. The main contents and innovations of this paper are summarized as follows:
(1) Delineation of macaque brain parcellation based on diffusion MRI. The
brain is an integrated system composed of connections, and the connections are closely associated with the function of brain regions. Therefore, based on high spatial and an-gular resolution macaque ex vivo diffusion magnetic resonance imaging data, this study extracted whole-brain structural connectivity and established an automated parcellation pipeline for the macaque brain. Based on the connectivity with the diffusion MRI, we delineated a macaque brain parcellation with fine-grained divisions. This atlas di-vided the macaque brain into 256 cortical and 58 subcortical subregions. Furthermore, through repeatability validation on a set of publicly available datasets, we confirmed the robustness and reproducibility of the delineation method and results. Additionally, comparative analysis with existing atlases revealed that the macaque brain constructed in this paper could more accurately represent the topology of brain connectivity. (2) Integration and validation of multimodal information with the connectivity-based macaque brain atlas. Existing macaque brain atlases often contain only one modality of information, which fails to comprehensively describe the brain. There-fore, based on the connectivity-based macaque brain atlas, this study incorporates mul-timodal MRI data to establish both structural and functional connections between eachbrain region. Subsequently, we integrate histological data and tracing data from the same macaque based on the the brain atlas, systematically validating the reliability of the macaque brain atlas. This atlas integrates multimodal information such as structural connections, tracing connections, and histology, thereby addressing the limitation of past macaque atlases in reflecting only a single modality profile of the brain. It provides an important avenue for comprehensive understanding of the macaque brain. (3) Comparative study of brain connectivity based on Human and Macaque Brainnetome atlases. During evolution, the connectivity patterns of brains in differ-ent species undergo changes. By comparing the differences in brain connectivity pat-terns between humans and macaques, we can better understand the unique higher-order functions of humans. Based on the Macaque and Human Brainnetome Atlases, this study initially analyzes the differential patterns of connectivity between the left and right hemispheres of macaque and human brains, respectively. Subsequently, using the Macaque and Human Brainnetome Atlases as comparative units, we describe the cross-species divergence in connectivity between macaque and human brains. Finally, through a joint analysis of connectivity differences and transcriptome data of human brain, we identify the gene sets that most likely drive these cross-species connectivity divergence during evolution. Through gene enrichment analysis, we discover that this gene set is closely related to human intelligence and psychiatric disorders. It suggests that the cross-species connectivity divergence may simultaneously contribute to the en-hancement of human intelligence and the emergence of psychiatric disorders. (4) Individualization of macaque brain atlases. In practical applications of macaque brain atlases, traditional methods often utilize morphology-based image reg-istration techniques to directly map the atlas onto individual brains, overlooking the individual differences in macaque brains that are similarly present in humans. To ad-dress this issue, this study proposes an individualization method of macaque parcella-tion based on brain connectivity, utilizing the group-level macaque brain atlas as a prior. This method mapped the individual feature with the group-averaged feature into a low-dimensional latent space and subsequently relabeled the macaque brain with Bayesian estimation. Compared with other individualization methods, the proposed approach can more accurately reflect the topology of individual connectivity.

Keyword脑图谱 猕猴 连接 弥散磁共振 细胞构筑 跨物种比较 个体化
Language中文
Document Type学位论文
Identifierhttp://ir.ia.ac.cn/handle/173211/57402
Collection脑图谱与类脑智能实验室_脑网络组研究
Recommended Citation
GB/T 7714
陆玉恒. 猕猴脑网络组图谱绘制及其应用[D],2024.
Files in This Item:
File Name/Size DocType Version Access License
Thesis_提交版20240604.p(45283KB)学位论文 开放获取CC BY-NC-SA
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[陆玉恒]'s Articles
Baidu academic
Similar articles in Baidu academic
[陆玉恒]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[陆玉恒]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.