ArticlePDF Available

人脑形态网络及其在脑发育研究中的应用

Authors:
Brain morphological network and its applications in human brain developmentBrain morphological network and its applications in human brain development
赵若可1, 赵智勇1, 王金辉2 and 吴丹1,*
Citation: 科学通报; doi: 10.1360/TB-2022-0621
View online: https://engine.scichina.com/doi/10.1360/TB-2022-0621
Published by the 《中国科学》杂志社
Articles you may be interested inArticles you may be interested in
Population normative models of human brain growth across development
Science Bulletin 6565, 1872 (2020);
Spatial multi-scaled chimera states of cerebral cortex network and its inherent structure-dynamics relationship in human brain
National Science Review 88, nwaa125 (2021);
Ketones, omega-3 fatty acids and the Yin-Yang balance in the brain: insights from infant development and Alzheimer’s disease, and
implications for human brain evolution
OCL (Oilseeds, Crops, fats & Lipids) 2525, D409 (2018);
The offer network protocol: Mathematical foundations and a roadmap for the development of a global brain
European Physical Journal ST(Special Topics) 226226, 283 (2017);
Transgenic rhesus monkeys carrying the human MCPH1 gene copies show human-like neoteny of brain development
National Science Review 66, 480 (2019);
人脑形态网络及其在脑发育研究中的应用
赵若可1,赵智勇1,王金辉2,吴丹1*
1. 浙江大学生物医学工程与仪器科学学院,杭州 310027;
2. 华南师范大学脑科学与康复医学研究院,广州 510631
*联系人, E-mail: danwu.bme@zju.edu.cn
2022-06-01 收稿, 2022-08-05 修回, 2022-08-08 接受, 2022-08-09 网络版发表
国家自然科学基金(61801424, 81971606, 82122032)国家重点研发计划(2018YFE0114600)浙江省领军型创新创业团队(202006140)和浙江
省重点研发计划(2022C03057)资助
摘要 人脑是一个极其复杂的系统,数以万亿的神经元通过突触构成了庞大的连接网络,如何系统而全面地刻画
大脑内部的组织模式一直是神经科学家致力解决的热点问题.随着非侵入神经成像技术的兴起,基于磁共振成像
的脑网络研究为揭示人脑的复杂结构提供了新的契机.其中,脑形态网络因其图像易获得质量稳定和分析方法
简单等优势引发了研究者的广泛关注.传统的脑形态网络是利用一组受试者的某种形态学指标构建的被试间协方
差网络,只能反映大脑形态的群组特征.个体水平脑形态网络则针对个体大脑刻画脑区间的形态相似性,保留了个
体的形态信息.多指标甚至多模态数据的应用,进一步整合多种互补的结构信息,能够综合表征皮层形态的拓扑连
接模式.近年来,脑形态网络在探究人类大脑发育进程方面体现了重要价值.研究发现,胎儿时期形态网络从高度
模块化的原始状态开始逐渐整合,至出生时已形成小世界拓扑,出生后至童年早期网络变得分离,童年晚期至青春
期又逐渐整合,初级网络率先达到成熟而高级网络长期持续发育.这些结论为理解大脑认知功能的形成和发育障
碍的起源提供了重要的理论支撑.本文介绍了两种脑形态网络的构建方法和图论模型的基本概念,回顾了脑形态
网络在胎儿婴儿儿童以及青少年大脑发育中的研究,总结了脑形态网络的发育规律,最后指出了该技术目前
存在的问题与面临的挑战.
关键词 磁共振成像,脑形态网络,拓扑组织,脑发育,分离与整合
人脑是一个极其复杂的系统,大量的神经元通过
突触互相连接,构成了庞大的解剖连接网络.这种网络
被认为是信息加工与认知表达的生理基础[1].随着非侵
入神经成像技术的兴起,人们对脑网络的探索进入了
一个新时代[2].研究者利用磁共振成像(magnetic reso-
nance imaging, MRI)描绘了人脑的解剖结构和功能同
,并结合基于图论的复杂网络分析方法,进一步揭示
了人脑的生理机制[1].
基于磁共振成像技术构建的脑网络可以分为脑形
态网络,脑白质纤维网络和脑功能网络3.脑形态网
络指利用结构磁共振成像(structural magnetic resonance
imaging, sMRI)构建的大脑形态连接网络;脑白质纤维
网络指利用弥散磁共振成像(diffusion magnetic reso-
nance imaging, dMRI)构建的脑白质连接网络;脑功能
网络指利用脑电图(electroencephalogram, EEG)脑磁
(magnetoencephalography, MEG)和功能磁共振成像
(functional MRI, fMRI)等构建的脑功能连接网络.
形态网络刻画了脑区间的形态学相似性;脑白质纤维
网络反映了脑区间的纤维连接情况;脑功能网络则描
述了不同脑区的协同工作模式.相比于脑白质纤维网
引用格式: 赵若可, 赵智勇, 王金辉, . 人脑形态网络及其在脑发育研究中的应用. 科学通报, 2022, 67
Zhao R K, Zhao Z Y, Wang J H, et al. Brain morphological network and its applications in human brain development (in Chinese). Chin Sci Bull, 2022, 67,
doi: 10.1360/TB-2022-0621
© 2022《中国科学》杂志社 www.scichina.com csb.scichina.com
https://engine.scichina.com/doi/10.1360/TB-2022-0621
络和脑功能网络,脑形态网络具有图像易获得应用
范围广分析方法相对简单生理意义相对明确等优
,近些年来已经涌现了许多相关研究.脑形态网络通
常是通过检查结构磁共振数据中脑区间的形态学特征
(如皮层厚度曲率表面积及沟回深度和灰质体积
)的统计相关性来获得的[3~7].其中,根据构建网络的
方式可以分为被试间协方差网络和个体水平脑形态网
.被试间协方差网络,顾名思义,是利用一组受试者
的形态学特征计算脑区间的协同变化而构建的网络;
个体水平脑形态网络则是针对一个受试者分析不同脑
区之间的形态相似性而构建的网络.
脑形态网络方法为研究人脑解剖结构认知功
脑区特异的基因表达和临床疾病的发病机制等提
供了新的视角[8~11].多项研究表明,脑形态网络与脑白
质纤维网络高度相似[12~14].有研究指出,皮层厚度呈正
相关的脑区之间大多数存在白质纤维束连接,而几乎
所有皮层厚度呈负相关的脑区之间都没有匹配的纤维
束连接,提示形态学连接部分反映了潜在的纤维束连
[12]. Seidlitz等人[10]通过将个体水平脑形态网络与经
典细胞结构图谱对齐,发现网络的拓扑结构与皮质区
域的细胞分类密切对应.这些发现说明,脑形态网络的
连通性可能由微观或宏观解剖结构驱动.同样地,脑形
态网络与脑功能网络也显示出相似的连接模式[11,15],
许多在形态学上高度相关的脑区同处于具有特定行为
或认知功能的系统[16~19].此外,多项研究致力于探索基
因表达与脑形态网络的关系.一项研究发现,5-
HTTLPR基因对杏仁核-前扣带回后部的灰质体积协方
差具有显著影响[20];个体水平脑形态网络的拓扑结构
也被证明与某些重要基因的空间表达模式一致[10].
而言之,脑形态网络特定的拓扑结构可能具有解剖和
遗传基础,与大脑的认知功能相关,并且能够为临床疾
病提供有效的生物学标记.目前,脑形态网络已应用于
探索临床疾病的发病机制中,阿尔茨海默病[21]精神
分裂症[22]癫痫[23]和自闭症[24]等多种疾病都表现出异
常的形态相似模式,这些发现对患者的临床诊断和康
复治疗具有一定的指导意义.
本文将聚焦脑形态网络的新近发展及其在脑发育
研究中的应用,首先从组水平和个体水平介绍形态网
络构建的多种方法以及图论模型的基本概念;然后综
述近年来脑形态网络随发育变化的研究,总结脑形态
网络的发育规律;最后,讨论目前该领域所存在的问题
与面临的挑战.
1脑形态网络的构建方法与图论的基本概念
1.1 脑形态网络的构建方法
脑形态网络的构建一般包含以下几步:首先,将高
分辨率的T1加权或T2加权图像配准到标准空间,并进
行偏置场校正和分割等预处理.然后,计算皮层厚度
曲率表面积和沟回深度等形态学指标,并根据已有
模板提取脑区内的平均值.最后,计算不同脑区之间形
态学指标的相关性,从而构建脑形态网络(1). 早期的
脑形态网络研究通常使用一种指标(皮层厚度皮层
曲率灰质体积灰质密度等)构建被试间协方差网
,这种网络只能在群体水平反映大脑形态结构的协
同变化模式,无法反映单个个体的形态特征.这限制了
其在研究个体大脑形态变异性,特别是识别患者大脑
形态异常方面的应用[27].后续研究提出了多种构建个
体水平脑形态网络的方法,这些方法不仅能量化脑区
的形态相似性,同时反映了个体大脑的形态信息,为个
体化研究提供了可能[7,31~33].下面对两类脑形态网络的
构建方法进行具体介绍.
1.1.1 被试间协方差网络
2005, Mechelli等人[34]根据结构磁共振图像发现
某些脑区的灰质密度具有协调的变化趋势. 2007, He
等人[3]采集了健康受试者的T1加权磁共振图像,通过
计算脑区皮层厚度之间的皮尔逊相关系数构建了被试
间协方差网络,并结合图论模型发现该网络具有小世
界属性.因被试间协方差网络具有数据易得分析方
法简单等优点,在国内外迅速掀起了研究热潮,近十余
年涌现了相当多的研究.多数研究延续He等人[3]提出
的思路,通过将不同个体相同脑区的皮层形态学指标
构成向量并计算特征向量的相关性,从而构建被试间
协方差网络,但在形态学指标的选择脑区的分割和
统计相关性的计算方法上存在差异.个体大脑的形态
学指标一般包括皮层厚度曲率面积及沟回深
[6,35]或灰质体积密度等[4,19],它们反映了不同的皮
质形态特性,因而会得到不同的网络拓扑[14].在脑区的
分割上,除了选择已有的脑图谱(AALDesikan-
KillianyANIMAL模板)[36~38],一些研究直接将感兴趣
区域作为种子点,计算其与大脑其他区域(通常是体素)
的结构共变模式[25,35].相较而言,统计相关性的计算方
法则比较单一,研究者通常计算皮尔逊相关系数或偏
相关系数[21,35]作为连接强度.如上所述,被试间协方差
网络具有显著的缺点,它只能为每一组受试者建立一
2 https://engine.scichina.com/doi/10.1360/TB-2022-0621
个网络,该网络反映了受试者的群组特征,而不能反映
个体大脑的形态信息[39],为个体化研究带来了困难.
1.1.2 个体水平脑形态网络
近些年,一些研究者提出了基于个体水平构建脑
形态网络的方法. Tijms等人[25]通过将大脑灰质分割成
大小相同的小立方体,并比较不同立方体之间灰质强
度的相似性,建立了个体水平的脑形态网络. Wang
[28]利用3D小波变换[40]将灰质体积图像转换到小波
,选择一系列主小波系数作为体素的形态特征向量,
从而构建基于体素的个体水平脑形态网络. Wee
[30]提出了一种基于差异性的边缘计算方法,通过计
算脑区之间皮层厚度的平均值和标准差的差异来表示
区域的相似性. Kong等人[26,27]则采用了信息论中相对
熵的概念,3D体积数据上估算了区域灰质强度分布
1(网络版彩色)个体水平脑形态网络和被试间协方差网络的构建与分析流程. (a) 结构磁共振和弥散磁共振等多模态图像; (b) 皮层形态学
指标和扩散参数; (c) 脑区分割; (d) 个体水平脑形态网络(基于单一特征的方法包括:基于立方体的方法[25];基于相对熵的方法[26~28];
于小波变换的方法[29];基于差异性的方法[30])被试间协方差网络; (e) 相关性矩阵; (f) 二值矩阵; (g) 图论分析. CT, 皮层厚度; SA, 表面积;
MC, 平均曲率; FA, 部分各向异性; MK, 平均峰度; ODI, 方向分散度
Figure 1 (Color online) Construction and analysis of individual-level and across-subjects covariance network. (a) Structural and diffusion magnetic
resonance images; (b) cortical morphological features and diffusion parameters; (c) segmentation of brain regions; (d) constructions of individual-level
brain morphological network (single feature based methods including: cube-based[25],KL divergence-based[26–28],wavelet transform-based[29],
dissimilarity-based[30], etc.), as well as across-subjects covariance network; (e) correlation matrix; (f) binarized matrix; (g) graph theory analysis. CT,
cortical thickness; SA, surface area; MC, mean curvature; FA, fraction anisotropic; MK, mean kurtosis; ODI, orientation dispersion index
3
https://engine.scichina.com/doi/10.1360/TB-2022-0621
的相似性,从而建立全脑的形态连接.有研究将Kong
[26,27]的方法进一步应用在2D表面数据上,通过KL
[41]JS散度[42]计算形态指标的概率密度函数的相似
性来量化区域间关系[43].同样,还有一些研究收集了个
体的纵向磁共振数据研究脑区形态的成熟轨迹,称其
为成熟耦合网络(subject-based maturational coupling
network, sbMCN)[44].
除了以上这些使用单一的形态学特征(如皮层厚度
或灰质体积)构建脑形态网络的方法,研究者还提出了
利用多种形态学特征描述皮层形态相似性的方法(
1), 不仅能够评估个体水平的形态变异性[10],也可实现
对大脑灰质形态组织模式的全面系统化表征.
基于多特征的个体水平脑形态网络依据成像模式
又可分为单模态网络和多模态网络.单模态网络通常
指利用结构磁共振图像数据获取大脑灰质的多种形态
学特征(皮层厚度表面积沟回深度平均曲率和
灰质体积等[10,45~49]), 计算不同脑区特征向量之间的相
似性来构建网络的方法.Li等人[45]通过计算7个形态
特征的皮尔逊相关系数构建脑形态网络,并发现该网
络具有小世界属性. Zhao等人[49]提取每个脑区的放射
组学特征来表征不同脑区的形态特点,从而建立脑区
间的形态相似关系. Yu等人[46]以多元欧几里得距离的
指数函数作为两个区域之间相似性的评估,构建基于
多特征的脑形态网络. Zheng等人[7]以目标区域的特征
向量作为目标变量,所有其他区域的特征向量作为回
归变量,使用稀疏线性回归(least absolute shrinkage
and selection operator, LASSO)来量化各种形态特征的
跨区域关系.
结构磁共振图像提供了大脑皮层的宏观结构信息,
弥散磁共振图像则可以提供组织的微观结构和神经纤
维走行等信息.基于多模态的脑形态网络正是利用这
些互补的信息综合刻画了脑区间的形态关系. Seidlitz
等人[10]将弥散磁共振图像和结构磁共振图像相结合,
利用皮层曲率髓鞘指数和灰质形态三类形态特征构
建基于多模态的脑形态网络,并发现该网络与大脑皮
质的细胞结构分类高度相似,更高的区域间连接强度
也与特定基因的共表达十分相关.除了传统的弥散加
权磁共振成像技术,研究者还将弥散张量成像(diffu-
sion tensor imaging, DTI)弥散峰度成像(diffusion
kurtosis imaging, DKI)神经突方向分散度和密度成
像技术(neurite orientation dispersion and density ima-
ging, NODDI)等多种模型参数加入到特征向量中,旨在
提供脑区形态特性的多维描述[50,51],使得基于多模态
的脑形态学网络在综合量化不同脑区的形态相似性上
更具优势.本文不涉及由弥散磁共振图像构建的白质
纤维网络,感兴趣的读者可以参考其他综述[52,53].
1.2 图论的基本概念
图论模型为复杂脑网络的分析提供了有力的工具.
在图论模型中,节点通常代表脑区,边代表脑区之间的
形态白质纤维或功能连接.根据连接是否具有方向
,可将网络分为有向和无向网络;根据连接强度的大
,可将网络分为加权和二值网络.有向加权网络面临
计算量大生理意义难以解释等问题.因此,多数研究
为了简化分析将脑网络抽象成无向二值网络,通过研
究其拓扑属性来揭示脑网络的组织原理和演化规律.
1列出了常用的无向二值网络的拓扑属性.
2脑形态网络在人脑发育研究中的应用
在人类发育过程中,大脑始终处于不断变化的状
.从胚胎到青春期,每个阶段的变化对认知语言
记忆情绪以及思维功能都至关重要.出生前,神经形
成基本上在受孕后20周完成,孕中后期轴突开始生长,
神经元之间的突触慢慢形成.出生后,突触数量进一步
增加并发生细胞损失突触修剪和髓鞘形成等过程,
一直持续到青春期[58,59].然而,这些微观层面的神经结
构改变如何影响大脑灰质形态的重塑,一直是近年来
脑发育领域的热点问题.脑形态网络为这些问题的研
究提供了一种思路.下面将分别介绍脑形态网络在胎
婴幼儿儿童和青少年大脑发育中的研究,最后
总结脑形态网络的发育规律.
2.1 胎儿与婴幼儿
在非侵入性神经成像技术的推动下,人们对胎儿
和婴儿大脑发育的探索进入了一个新时代,通过分析
脑形态网络拓扑属性的变化,可以探究出生前后大脑
形态结构是如何重组的. Fan等人[4]采集了28名婴幼儿
出生后1个月1岁和2岁时的MRI图像和27名成年人
MRI图像,利用灰质体积构建了被试间协方差网络.
结果发现, 1个月婴儿的脑形态网络即具有典型的小世
界拓扑属性和非随机的模块化分布,网络的全局效率
随着发育呈现增长模式,2岁幼儿脑形态网络的模块
化程度和局部效率均高于年幼儿童与成人大脑中模块
化程度和局部效率. Nie等人[14]则发现,从出生至2,
4 https://engine.scichina.com/doi/10.1360/TB-2022-0621
1无向二值网络的拓扑属性[54~56]
Table 1 Topological properties of undirected binary networks[54–56]
属性 描述 数学定义
基本属性
(degree) 节点的直接连边数,是对节点连接强度的最直接
描述,也反映了重要的网络演化特性[54]
节点i的度ki定义为
k a= ,
ij N ij
其中,N是网络中节点的集合,aij表示节点i和节点j的连接状态,aij=1
表示连接存在,aij=0表示连接不存在
最短路径长度
(shortest path length)
指一个节点到另一个节点所需经过的最少边数,
度量了节点之间的信息传输速度
节点i到节点j的最短路径长度dij定义为
d a= ,
ij auv gi j
uv
其中,
gi j
是节点i和节点j之间的最短路径,uv是最短路径上的节点
网络整合
特征路径长度(charac-
teristic path length)
网络中所有节点对的平均最短路径长度,量化了
网络的全局传输能力
网络的特征路径长度L定义为
LnLn
d
n
=1=1
1,
i N ii N
j N j i ij
,
其中,n是网络中节点的数量,Li是节点i的特征路径长度
全局效率(global effi-
ciency)
网络中所有节点对的最短路径长度倒数的平均.
相比于特征路径长度,全局效率更适用于描述包
含孤立节点的网络的整合程度[55]
网络的全局效率E定义为
EnEn
d
n
=1=1
1,
i N ii N
j N j i ij
,
1
其中,Ei是节点i的效率
网络分离
集群系数(clustering
coefficient)
节点的邻居节点间的实际连边数与可能的最大
连边数的比值,衡量了网络的集团化程度.网络
的集群系数是网络所有节点的集群系数的平均
网络的集群系数C定义为
其中,Ci是节点i的集群系数
局部效率(local
efficiency)
节点的邻居节点构成的子图的效率.网络的局部
效率是网络所有节点的局部效率的平均
网络的局部效率Eloc定义为
( )
EnEn
a a d N
k k
=1=1
( 1) ,
i N ii N
j h N j i ij ih jh i
i i
loc loc,
, ,
1
其中,Eloc,i表示节点i的局部效率
模块化(modularity) 用于评价网络社区划分的优劣[57]
网络的模块化Q定义为
Q e e= ,
u M uu v M uv
2
其中网络被完全分割为一组不重叠的模块M,euv是模块u与模块v的连接
占网络中所有连接的比例
中心性
中介中心性
(betweenness centrality)
网络中所有节点对的最短路径通过该节点的比
,从信息流的角度衡量节点的中心程度
节点i的中介中心性bi定义为
bn n
i
=1
( 1)( 2)
( )
,
i
h j N
h j h i j i
hj
hj
,
, ,
其中,
hj
是节点hj之间最短路径的数目,
i( )
hj
是节点hj最短路径中
通过节点i的数目
接近中心性(closeness
centrality)
节点到所有其他节点的平均最短路径长度的倒
,反映了节点与网络中其他节点的接近程度
节点i的接近中心性
Li
1
定义为
Ln
d
=1
i
j N j i ij
1
,
其他属性
小世界性(small-
worldness)
小世界性是用于量化网络是否属于小世界网络
的指标
网络的小世界性S定义为
CC
LL
S= ,
rand
rand
其中,CL是真实网络的集群系数和特征路径长度,CrandLrand
随机网络的集群系数和特征路径长度
富人俱乐部系数
(“rich-club” coefficient)
网络中的核心节点(hub)比其他节点更为紧密连
接的现象.对于网络中的每一个节点度(k), 富人
俱乐部系数定义为,所有节点度大于k的节点构
成子图的实际连边数与所有可能连边数的比值
网络在节点度k下的富人俱乐部系数
k( )
定义为
k
E
N N
( ) = 2
( 1) ,
k
k k
>
> >
其中,
Nk>
表示度大于k的节点的数目,
Ek>
表示度大于k的节点构
成子图的实际连边数
5
https://engine.scichina.com/doi/10.1360/TB-2022-0621
基于皮层厚度的被试间协方差网络的全局效率几乎不
,而局部效率降低,相反,基于曲度的网络局部效率
增加,全局效率降低.另外,基于皮层厚度网络的中心
节点在出生后逐渐从高级功能区(额叶)消失,基于曲度
网络的中心节点则始终保持在不断发育的高级功能区
(额叶)并从早期成熟区域(中央后回)消失. Geng等人[15]
基于皮层厚度构建被试间协方差网络,研究了婴幼儿
大脑形态网络在2岁以内的变化,并将其与功能网络进
行了比较.结果发现, 2岁时基于默认和背侧注意种子
点构建的脑形态网络的分布比其功能网络的分布小许
,初级视觉感觉运动网络的脑形态网络显示出与
相应功能网络相似的模式.最近一项研究采集了婴儿
期到儿童期(0~6)的纵向结构磁共振数据,利用皮层
厚度构建不同年龄下的被试间协方差网络,同时利用
个体的纵向数据构建个体水平成熟耦合网络.结果发
,被试间协方差网络的模块化程度随年龄增加,全局
效率随年龄降低.并且,早期发育(0~1)时更低的个体
水平成熟耦合网络全局效率与8~10岁时运动学习的优
异表现有关[60]. Fenchel等人[50]发现,新生儿的多模态个
体水平脑形态网络表现出明显的模块化分布,该分布
模式在空间上与已知的功能网络和细胞结构类别相对
;随着年龄的增长,模块内的相似性逐渐增加,提示
模块间的差异逐渐显现.此外,新生儿的多特征脑形态
网络不仅可以预测婴儿时期的社会情绪表现[61],与使
用单一指标相比,多特征脑形态网络在区分早产婴儿
和足月婴儿上也表现出更好的性能[51].
以上研究均关注脑形态网络在婴幼儿发育中的变
,因子宫内胎儿结构磁共振成像面临许多困难,包括
胎儿运动母体呼吸产生的伪影以及图像重建困难等,
目前有关胎儿(产前)脑形态网络的研究非常少.最近,
Xu等人[62]采集14~22孕周胎儿的离体大脑结构磁共振
图像,构建被试间协方差网络.他们发现,相比新生儿,
孕中期网络在不同的节点度和最短路径长度下的节点
数目大致相似,并且网络连接以短距离连接为主.这表
,孕中期前期胎儿的脑形态网络的拓扑结构和网络
属性呈现不成熟和低效的特点,整合程度低但分离程
度高[62];然而,随着孕中晚期大脑皮层经历复杂的表面
扩张和折叠过程[63],脑形态网络的拓扑结构发生重组,
网络中长距离连接的出现使得部分节点进化成为核心
节点,网络的小世界特性逐渐显现[62].由于离体大脑成
像需要复杂的样本处理过程,大脑皮层形态可能发生
不同程度的形变,因此基于宫内胎儿的活体大脑成像
对产前大脑形态学研究至关重要.随着前沿成像技术
和后处理技术[64~66]的迅速发展,我们预期这一目标将
在不久的将来成为可能.
2总结了脑形态网络在胎儿和婴幼儿大脑发育
中的研究.总的来说,人类大脑皮层形态的拓扑结构在
孕中期前期仍处于不成熟高成本的状态.孕中后期
皮层形态快速变化[63],脑形态网络发生重塑,网络的整
合程度升高并逐渐向低成本高效率的结构转变.
生时,人脑形态网络已具有小世界拓扑和模块化组织,
这些模块化组织可以部分映射到皮层的细胞架构与功
能分区上.出生后,不同的形态学指标刻画了皮层的不
同发育轨迹,婴幼儿脑形态网络总体偏向更分离的方
向发展;皮层网络进一步分化,初级网络(初级视觉和
感觉运动网络)随年龄增长快速发育, 2岁时其形态连
接已经具有与相应功能网络相似的模式,而默认模式
和背侧注意网络等高级网络的形态连接尚未完全形成.
2.2 儿童和青春期
2010, Zielinski等人[19]首次将基于灰质密度的被
试间协方差网络应用于儿童与青少年的发育研究中,
发现初级感觉运动皮层网络在童年早期就已经发育良
,并在青春期前期继续扩张并修剪成类似于成人连
接模式的拓扑结构.相比之下,语言和社会情感等高级
认知网络在年幼群体中尚未发育完全,在年长儿童中
表现出越来越分散的拓扑结构. 2013, Khundrakpam
等人[35]进一步采用皮层厚度来检查儿童和青少年被试
间协方差网络的发育变化,发现从童年早期进入到童
年晚期后,网络的局部效率模块化降低,而全局效率
提高,提示该时期网络的拓扑结构向更随机的分布转
.进入青春期后,网络的全局效率略有下降但局部效
率略有上升,后基本保持不变;初级感觉运动区在童年
早期就已经成熟,而联合皮质区和边缘旁系区域在童
年和青春期持续发育. Nie等人[6]则采集了3~20岁健康
受试者的纵向结构磁共振图像,发现被试间协方差网
络的发育在7岁左右达到一个转折点,全局和局部效率
分别达到最低和最高值;网络的变化在这个年龄也达
到最高水平. Whitaker等人[71]利用皮层厚度构建了
14~24岁受试者的被试间协方差网络,发现网络是一个
具有模块结构和富人俱乐部的小世界网络,网络的
核心节点主要位于额顶叶联合皮质区. Váša等人[73]
Whitaker等人[71]研究的启发,以滑动窗的方法构建了
14~24岁受试者在不同年龄下的被试间协方差网络,
6 https://engine.scichina.com/doi/10.1360/TB-2022-0621
现网络强度随着发育降低;联合皮质区的结节度在14
岁时开始急剧下降,约在19岁时达到最低值,之后趋于
平稳或略有增加;青春期迅速变化的脑区主要集中在
前额叶;7种功能网络[72],额顶叶控制网络内部以
及与其他网络之间的形态连接随年龄下降最多. Khun-
drakpam等人[44]采集正常发育儿童和青少年的纵向磁
2胎儿婴幼儿的脑形态网络研究a)
Table 2 Brain morphological networks in fetuses and infants
作者 年份 被试年龄(数量)节点 指标(网络)主要发现
Fan等人[4] 2011
1个月(28)
1(28)
2(28)
24±3(27)
90(AAL图谱)灰质体积(被试间协
方差网络)
1个月时脑形态网络即具有典型的
小世界拓扑和非随机的模块化组
.网络的全局效率随发育呈现增
长模式. 2岁幼儿大脑形态网络的模
块化程度和局部效率均高于年幼儿
童与成人大脑中模块化程度和局部
效率
Nie等人[14] 2014
新生儿(73)
1(73)
2(73)
78(AAL图谱)
皮层厚度(被试间协方差
网络)曲度(被试间协方
差网络)纤维密度(脑白
质纤维网络)
基于皮层厚度的脑形态网络在出生
2年内局部效率降低,网络的核心
节点从额叶消失,在视觉区和中央
后回出现.基于曲度的脑形态网络
全局效率下降,局部效率上升,核心
节点从中央后回消失,在额叶中保
持不变.脑白质纤维网络的拓扑结
构和核心节点在12岁时保持一致
Geng等人[15] 2017
新生儿(118)
1(100)
2(82)
体素 皮层厚度(被试间协
方差网络)功能网络
2岁时,基于默认和背侧注意功能网
络种子点的被试间协方差网络的分
布比其功能网络的分布小许多,
于初级视觉和感觉运动功能网络种
子点的网络显示出与相应功能网络
相似的模式
Galdi等人[51] 2020 早产儿(59)
足月儿(46) 81(Draw-EM[67]图谱)
灰质体积+
T1w/T2w+
5DK[69]模型指标+
5NODDI[70]模型指标(
模态个体水平脑形态网络)
网络中编码的信息可用于预测围产
期大脑的实际年龄,并且提供了一
种新的数据驱动的方法,用于研究
与早产相关的神经解剖学变异
Fenchel等人[50] 2020 37~44(241) 150(平均分割)
皮层厚度+
皮层表面积+
T1w/T2w+
平均曲率+
2DTI[68]指标+
2NODDI指标(多模态
个体水平脑形态网络)
网络的模块化组织在空间上与已知
的功能网络和细胞结构类别相对
.随着年龄的增长,后部区域在形
态上变得更加相似,而扣带回周围
和内侧颞叶则变得更加不同.模块
内的节点相似性随着年龄增长而
增加
Woodburn
[60] 2021
0/3/6/9个月(74/58/60/50)
12/18/24/36个月(49/42/26/16)
48/60/72个月(29/28/20)
146(Destrieux图谱)
皮层厚度(被试间协
方差网络个体水平
成熟耦合网络)
从婴儿早期到6岁之间,被试间协方
差网络的模块化程度增加全局效
率降低.并且在早期发育阶段(0~1
), 个体水平成熟耦合网络的较低
全局效率可能与8~10岁时更优异的
运动学习表现有关
Xu等人[62] 2021 14~22周离体胎儿(27)
38~42周新生儿(40) 1284(体素)皮层厚度(被试间协
方差网络)
孕中期前期胎儿的脑形态网络的拓
扑结构和网络属性呈现出不成熟和
低效的特点,整合程度低但分离程
度高
Fenchel等人[61] 2022 37~44(193) 143(平均分割)
皮层厚度+
皮层表面积+
T1w/T2w+
平均曲率+
2DTI指标+
2NODDI指标(多模态
个体水平脑形态网络)
新生儿个体脑形态网络成功预测了
18个月时的社会情绪表现.预测边
缘主要分布在初级和后部皮层区域
a) AAL, 一种自动解剖标记图谱; Draw-EM, 一种基于期望最大化的发育大脑区域注释算法
7
https://engine.scichina.com/doi/10.1360/TB-2022-0621
共振图像,根据皮层厚度构建基于个体的成熟耦合网
.他们发现,女性同源脑区之间的成熟耦合连接更
,默认模式网络中成熟耦合的发育规律与被试间协
方差网络和功能连接的发育变化相一致. Vijayakumar
等人[74]8~14岁大脑的被试间协方差网络进行研究,
发现平均连接强度和网络密度遵循非线性的变化规律,
11岁左右达到峰值;在童年晚期,感觉运动皮层作为高
度连接的核心节点出现,而联合皮质区的核心节点在
青春期中期增加.另外,背侧注意额顶网络的网络内
连接强度随年龄增加,视觉躯体运动网络的网络内
连接强度随年龄下降,腹侧注意背侧注意额顶和
默认模式网络之间的连接强度随年龄增加.
3总结了脑形态网络在儿童和青春期大脑发育
的相关研究.研究表明,童年早期网络的全局效率降低
而局部效率升高,初级感觉运动网络的形态连接已经
发育良好,语言和社会情感等高级网络的形态连接尚
未成熟;童年晚期全局效率升高而局部效率模块化
降低,网络的核心节点出现在感觉运动皮层;进入青春
期后,网络效率基本保持稳定,网络连接强度下降,
络的核心节点逐渐移动至联合皮质区,迅速变化的脑
区主要集中在前额叶,负责复杂认知功能的高级网络
的形态连接逐渐成熟.整体来说,脑形态网络的发育呈
现空间特异性,初级网络率先达到成熟而高级网络持
续发育.
2.3 脑形态网络的发育规律
以上研究表明,脑形态网络在发育过程中遵循先
逐渐整合然后变得分离又逐渐整合的变化规律(2),
产前的网络整合使得分散的模块间建立连接,随后的
网络分离加强了模块内部的连接,最后的网络整合进
一步加强了模块间连接.在孕中期早期,高成本低效
率的区域通信意味着脑形态网络仍处于高度模块化的
原始状态,小世界网络架构尚未完全建立.随着发育,
少量的长距离连接出现,网络全局整合能力增强,模块
化减弱[62].出生时,脑形态网络已具有小世界拓扑和模
块化的分布特点[4,50,62].从婴儿时期到童年早期,
块内部的连接强度增加,局部信息传递能力增强,
形态网络逐渐分离[4,6,60].其中,初级网络(视觉
觉和运动网络)快速发育并在童年早期率先达到成熟,
而高级网络(额顶注意默认模式网络等)持续发
[15,19,35,74].从童年后期到青春期,网络中的长程连接
增强,远距离节点间的信息传输变得更加高效,全局整
2(网络版彩色)脑形态网络随发育变化的甘特图(a)和拓扑示意图(b). (a) 彩条的强度表示网络分离整合的程度及初级高级网络的成熟
程度.三角形表示小世界特性的出现.脑形态网络在发育过程中遵循先逐渐整合然后变得分离又逐渐整合的变化规律,初级网络率先达到成熟
而高级网络持续发育. (b) 出生前网络拓扑处于高度模块化的原始状态.随着产前的发育,长程连接出现,网络的整合程度增加.出生时,网络已
经具有典型的小世界拓扑.出生后至童年早期,模块内部连接增强,网络的分离程度增加.童年后期至青春期,模块间的连接增强,网络的全局信
息传递更加高效
Figure 2 (Color online) Gantt chart (a) and topological diagram (b) of brain morphological network with development. (a) The intensity of color-bars
illustrates the degree of network separation and integration, and the maturity of the primary and higher-order networks. The triangle indicates the
appearance of small-worldness. Brain morphological network follows the pattern from gradually integrated to separated and then gradually integrated
during development. The primary networks first reach maturity, while the higher-order networks continue to develop for a long time. (b) Before birth,
the network topology is in its primitive state with high modularity. During prenatal development, long-range connections appear and network becomes
more integrated. At birth, the network already has the characteristic small-world topology. From birth to early childhood, the intra-module connection
strength increases and network becomes more segregated. From late childhood to adolescence, the inter-module communication is further enhanced, and
the transmission of global information in the network becomes more efficient
8 https://engine.scichina.com/doi/10.1360/TB-2022-0621
合程度提高,拓扑结构逐渐向成人模式靠近[6],初级网
络在相对成熟的状态下进一步完善,高级网络则开始
快速发育并达到成熟[19,35,74].
3问题与挑战
尽管脑形态网络经历了十几年的发展,但是这一
领域仍存在许多亟待解决的问题.
(1) 脑形态网络的构建.与其他形式的脑网络一样,
脑形态网络中的节点和边存在多种定义.节点的定义
可以分为基于顶点(体素)和感兴趣区域(region of inter-
est, ROI)两种方式.基于顶点的方法需要昂贵的计算成
,包含大量的冗余信息,并且可能高估同一区域内顶
3儿童和青春期的脑形态网络研究a)
Table 3 Brain morphological networks in children and adolescents
作者 年份 被试年龄(数量)节点 指标(网络)主要发现
Zielinski等人[19] 2010
4.9~7.8(75)
8.5~11.2(75)
12.1~14.2(75)
16~18(75)
体素 灰质强度(被试间协
方差网络)
初级感觉运动区域的脑形态网络在童年早
期已经发育良好并在青春期前期继续完善.
社会情感和其他认知区域的脑形态网络
在年幼儿童中相对不成熟,在年长儿童中表现
为越来越分散的拓扑结构
Khundrakpam等人[35] 2013
4.8~8.4(51)
8.5~11.3(51)
11.4~14.7(51)
14.8~18.3(50)
78(AAL图谱)皮层厚度(被试间协
方差网络)
童年晚期的局部效率模块化降低,全局效率
升高,核心节点的数量和分布范围增加.初级
感觉运动区较早成熟,高阶的联合皮质和边缘
旁系区域持续发育
Alexander-Bloch
等人[11] 2013 9.0~22.8(108) 360(平均分割)皮层厚度(被试间协
方差网络), 功能网络
脑形态网络和功能网络都具有非随机的小世
界和模块化拓扑结构.形态连接概括了协调的
皮层成熟模式和部分的功能激活模式
Nie等人[6] 2013
3/4/5(45/15/15)
6/7/8(35/50/75)
9/10/11(75/95/75)
12/13/14(75/60/60)
15/16/17(55/65/55)
18/19/20(40/20/20)
78(AAL图谱)
皮层厚度(被试间协
方差网络), 曲度(
试间协方差网络)
皮层的发育在7岁左右达到转折点:皮质厚度
达到最高值,皮质折叠变得稳定;皮层厚度
曲度脑形态网络的全局效率最低,局部效率最
,网络变化最明显
Whitaker等人[71] 2016
14~15(60)
16~17(60)
18~19(60)
20~21(60)
22~24(60)
308(平均分割)皮层厚度(被试间协
方差网络)
脑形态网络是一个具有模块结构和富人俱乐
的小世界网络,网络中的核心节点主要位
于额顶叶联合皮质区
Moura等人[13] 2017
7~8(83)
9~11(83)
12~14(83)
20(DK图谱)
皮层厚度(被试间协方
差网络), 各向异性分
(脑白质纤维网络)
7~14,大脑皮层随着年龄增长而变薄,伴随
着白质的各向异性分数的增加.脑形态和脑白
质纤维网络之间高度相似
Váša等人[73] 2018
14~15(~60)
16~17(~60)
18~19(~60)
20~21(~60)
22~24(~60)
308(平均分割)皮层厚度(被试间协
方差网络)
网络连接强度随着年龄而降低.14~24,
合皮质区的中心度先急剧下降后趋于平稳或
略有增加,大约19岁时达到最低值.网络中随
年龄变化最快的区域主要集中在前额叶模块,
额顶叶控制网络内部以及与其他网络之间的
形态连接随年龄下降最多
Khundrakpam等人[44] 2019
11.9±3.6
13.8±3.7
15.8±3.6(141)
78(AAL图谱)
皮层厚度(被试间协
方差网络,个体水平
成熟耦合网络)
同源和邻近的灰质区域具有更强的成熟耦合.
与男性相比,女性同源区域之间的成熟耦合更
.默认模式网络内成熟耦合的发育变化与其
形态连接和功能连接的发育变化一致
Vijayakumar等人[74] 2021 8.5~14.5(366) 360(HCP-MMP1[75]
图谱)
皮层厚度(被试间
协方差网络)
网络的连接强度和网络密度遵循着非线性的
变化规律.感觉运动皮层的核心节点在儿童晚
期出现,而联合皮质区作为核心节点的范围和
强度在青春期中期增加.背侧注意额顶网络
的网络内连接强度随年龄增加,视觉躯体运
动网络的网络内连接强度随年龄下降,腹侧注
背侧注意额顶和默认模式网络之间的
连接强度随年龄增加
a) DK, Desikan-Killiany图谱; HCP-MMP1, 人类连接组计划多模态分割1.0
9
https://engine.scichina.com/doi/10.1360/TB-2022-0621
点之间的连通性.大脑尺寸在发育过程中快速变化,
同的大脑尺寸会导致不同的节点数量,节点数量又会
明显地影响网络属性[76],因此基于顶点(体素)的方法
不适用于发育人群的研究.基于ROI的方法需要先验知
识对大脑进行分割,过粗的分割会导致连接模式混杂,
丢失特定的连接信息[77].值得注意的是,在婴幼儿数据
上采用成人脑模板进行分割可能会引起偏差.尽管目
前已经存在特定的婴幼儿脑模板[67,78,79],但这些模板对
大脑的分区不尽相同,其年龄范围也尚未涵盖全部发
育人群.因此,要想建立准确的形态学网络发育模型,
还需要更多精细年龄范围下的脑模板.同样地,网络中
边的定义也需要考虑类似的问题.在脑形态网络中,
通常定义为不同节点形态学度量(或它们的组合)的统
计相关性.形态学度量可以选择皮层厚度曲率和表
面积等指标,统计相关性也可以基于皮尔森相关
相关和斯皮尔曼相关等方法计算.不同的形态学度量
反映了脑区间的不同共变模式,不同相关性计算方法
也会引起网络属性的差异[80].总之,如何为节点和边确
定合适的定义方法以及如何对具有不同节点和边定义
的网络进行拓扑属性的定量比较也是今后的研究重点
之一.
(2) 网络分析.在进行网络分析之前,许多研究通
过选取适当的阈值获得无向二值网络.最初的二值化
方法是通过设置单一阈值来保留网络中权重更大的边,
以减少弱相关或虚假连接的影响.这种简单的二值化
方式可能会给网络拓扑属性的分析带来限制.目前常
用的方法是使用稀疏度为每个网络保留固定比例的最
强连接,以保证网络间的连接密度相同.多种稀疏度也
允许我们在一系列网络密度下探究小世界属性的变化
规律.此外,尽管这种无向二值网络分析简单计算量
,但它忽视了节点之间的方向和权重信息.构建有向
加权网络可以保留边的方向信息,便于深入刻画脑区
间的形态相似和因果关系. Zhang等人[81,82]提出了一种
利用格兰杰因果分析(granger causality analysis, GCA)
探究脑区间形态变化的因果关系方法,未来还需要进
一步研究这一问题.
(3) 脑形态网络的解释.一些研究通过对比脑白
质纤维网络与脑形态网络,认为形态连接部分反映了
白质纤维束的连接[12~14],但这样的相关性并不意味着
直接的因果关系,形态协方差背后可能存在着复杂的
解剖机制.与白质纤维连接相比,功能连接似乎更接
近于形态连接[9],在没有直接纤维束连接的情况下,
两个形态相似的脑区可能存在显著的功能连接[19,83].
研究发现,功能网络在形态网络之前就已经存在,
发育过程中两个具有相同功能的脑区在形态上也逐
渐变得相似[15].另外,大脑皮层的结构和功能均显示
出层次组织[48,84~86],通常称之为梯度.一些研究发现,
从童年到青少年的发育过程中,功能梯度逐渐从躯体
感觉运动皮层和视觉皮层等单模态皮层转移到默认
模式网络等联合皮层,揭示了皮层功能组织的特定成
熟模式[87,88].未来将梯度方法与形态学连接相结合可
以为理解大脑皮层的发育模式提供更多的参考.人们
也试图从基因和微观结构的角度寻求一些解释.研究
发现,转录组脑网络与脑形态网络有部分重叠的拓扑
结构,人类超颗粒富集(human supragranular enriched,
HSE)基因在网络核心节点中高度表达,脑区的基因
共表达也似乎证明了形态网络的模块化属性[10,71,89].
另外,微观解剖结构的重塑也是相当多研究推测的主
.一些研究证明,灰质形态协方差通常存在于具有
相同成熟轨迹的区域[11,15,44].这种同步成熟被认为与
协调的神经发育相关,如突触发生神经元迁移
鞘形成等.尽管已获得这些初步见解,我们仍然需要
对脑形态网络的生理意义进行更清晰而全面的解释.
尤其在多模态个体水平形态网络逐渐兴起之后,如何
对这种复杂的连接模式进行解释也需要进一步的
探索.
(4) 大脑发育早期成像的挑战.对早期形态结构进
行准确的网络建模,需要高质量的神经影像数据.对发
育早期的胎儿和婴幼儿进行结构磁共振图像采集时,
不可控的婴幼儿运动常常降低图像的质量,母体呼吸
也给体内胎儿成像带来了一定的伪影.与成人研究相
,胎儿和婴幼儿因大脑尺寸较小,使用常规成像序列
得到的灰白质之间组织对比度差,在图像预处理和分
割时也比成人更加困难[8].因此,如何缩短胎儿和婴幼
儿的成像时间,同时提高图像的分辨率和对比度[90],
优化图像处理和分割技术[91],以获取高质量的神经影
像数据成为该领域的一个重要课题.
4总结
综上所述,基于磁共振成像技术的脑形态网络使
我们能够从图论的角度分析和理解大脑灰质的拓扑结
,全面而细致地刻画大脑皮层的组织模式,从而可以
为神经发育行为认知的生理基础提供新的启示.
究表明,脑形态网络在发育过程中遵循先逐渐整合然
10 https://engine.scichina.com/doi/10.1360/TB-2022-0621
后变得分离又逐渐整合的变化规律.胎儿时期脑形态
网络从高度模块化的原始状态开始逐渐整合,至出生
时已形成小世界拓扑,出生后至童年早期网络变得分
,童年晚期至青春期网络又逐渐整合,初级网络率先
达到成熟,而高级网络长期持续发育.在过去的十几年
,被试间协方差网络方法已经趋于成熟,个体水平脑
形态网络正在快速发展,形态网络正逐渐从单模态单
特征向多模态多特征的方向演化.未来需要更多方法
学上的工作解决目前该领域存在的挑战,进一步推广
形态网络在大脑发育中的应用.
参考文献
1 Bullmore E, Sporns O. Complex brain networks: Graph theoretical analysis of structural and functional systems. Nat Rev Neurosci, 2009, 10: 186–
198
2 Zhao T D, Xu Y H, He Y. Graph theoretical modeling of baby brain networks. NeuroImage, 2019, 185: 711–727
3 He Y, Chen Z J, Evans A C. Small-world anatomical networks in the human brain revealed by cortical thickness from MRI. Cerebral Cortex, 2007,
17: 2407–2419
4 Fan Y, Shi F, Smith J K, et al. Brain anatomical networks in early human brain development. NeuroImage, 2011, 54: 1862–1871
5 Sanabria-Diaz G, Melie-García L, Iturria-Medina Y, et al. Surface area and cortical thickness descriptors reveal different attributes of the structural
human brain networks. NeuroImage, 2010, 50: 1497–1510
6 Nie J X, Li G, Shen D G. Development of cortical anatomical properties from early childhood to early adulthood. NeuroImage, 2013, 76: 216–224
7 Zheng W H, Yao Z J, Xie Y W, et al. Identification of Alzheimer’s disease and mild cognitive impairment using networks constructed based on
multiple morphological brain features. Biol Psychiat-Cogn Neurosci Neuroimag, 2018, 3: 887–897
8 Gilmore J H, Knickmeyer R C, Gao W. Imaging structural and functional brain development in early childhood. Nat Rev Neurosci, 2018, 19: 123–
137
9 Evans A C. Networks of anatomical covariance. NeuroImage, 2013, 80: 489–504
10 Seidlitz J, Váša F, Shinn M, et al. Morphometric similarity networks detect microscale cortical organization and predict inter-individual cognitive
variation. Neuron, 2018, 97: 231–247.e7
11 Alexander-Bloch A, Raznahan A, Bullmore E, et al. The convergence of maturational change and structural covariance in human cortical networks.
J Neurosci, 2013, 33: 2889–2899
12 Gong G L, He Y, Chen Z J, et al. Convergence and divergence of thickness correlations with diffusion connections across the human cerebral
cortex. NeuroImage, 2012, 59: 1239–1248
13 Moura L M, Crossley N A, Zugman A, et al. Coordinated brain development: Exploring the synchrony between changes in grey and white matter
during childhood maturation. Brain Imag Behav, 2017, 11: 808–817
14 Nie J X, Li G, Wang L, et al. Longitudinal development of cortical thickness, folding, and fiber density networks in the first 2 years of life. Hum
Brain Mapp, 2014, 35: 3726–3737
15 Geng X J, Li G, Lu Z H, et al. Structural and maturational covariance in early childhood brain development. Cereb Cortex, 2017, 27: 1795–1807
16 Alexander-Bloch A, Giedd J N, Bullmore E. Imaging structural co-variance between human brain regions. Nat Rev Neurosci, 2013, 14: 322–336
17 Bohbot V D, Lerch J, Thorndycraft B, et al. Gray matter differences correlate with spontaneous strategies in a human virtual navigation task. J
Neurosci, 2007, 27: 10078–10083
18 Lerch J P, Worsley K, Shaw W P, et al. Mapping anatomical correlations across cerebral cortex (macacc) using cortical thickness from MRI.
NeuroImage, 2006, 31: 993–1003
19 Zielinski B A, Gennatas E D, Zhou J, et al. Network-level structural covariance in the developing brain. Proc Natl Acad Sci USA, 2010, 107:
18191–18196
20 Pezawas L, Meyer-Lindenberg A, Drabant E M, et al. 5-HTTLPR polymorphism impacts human cingulate-amygdala interactions: A genetic
susceptibility mechanism for depression. Nat Neurosci, 2005, 8: 828–834
21 He Y, Chen Z, Evans A. Structural insights into aberrant topological patterns of large-scale cortical networks in Alzheimer’s disease. J Neurosci,
2008, 28: 4756–4766
22 Zhang Y C, Lin L, Lin C P, et al. Abnormal topological organization of structural brain networks in schizophrenia. Schizophr Res, 2012, 141: 109–
118
23 Bernhardt B C, Worsley K J, Besson P, et al. Mapping limbic network organization in temporal lobe epilepsy using morphometric correlations:
Insights on the relation between mesiotemporal connectivity and cortical atrophy. NeuroImage, 2008, 42: 515–524
24 Dziobek I, Bahnemann M, Convit A, et al. The role of the fusiform-amygdala system in the pathophysiology of autism. Arch Gen Psychiatry, 2010,
11
https://engine.scichina.com/doi/10.1360/TB-2022-0621
67: 397–405
25 Tijms B M, Seriès P, Willshaw D J, et al. Similarity-based extraction of individual networks from gray matter MRI scans. Cerebral Cortex, 2012,
22: 1530–1541
26 Kong X Z, Wang X, Huang L J, et al. Measuring individual morphological relationship of cortical regions. J Neurosci Methods, 2014, 237: 103–
107
27 Kong X Z, Liu Z G, Huang L J, et al. Mapping individual brain networks using statistical similarity in regional morphology from MRI. PLoS One,
2015, 10: e0141840
28 Wang H, Jin X Q, Zhang Y, et al. Single-subject morphological brain networks: Connectivity mapping, topological characterization and test-retest
reliability. Brain Behav, 2016, 6: e00448
29 Wang X H, Jiao Y, Li L H. Mapping individual voxel-wise morphological connectivity using wavelet transform of voxel-based morphology. PLoS
One, 2018, 13: e0201243
30 Wee C Y, Yap P T, Shen D. Prediction of Alzheimer’s disease and mild cognitive impairment using cortical morphological patterns. Hum Brain
Mapp, 2013, 34: 3411–3425
31 Morgan S E, Seidlitz J, Whitaker K J, et al. Cortical patterning of abnormal morphometric similarity in psychosis is associated with brain
expression of schizophrenia-related genes. Proc Natl Acad Sci USA, 2019, 116: 9604–9609
32 Seidlitz J, Nadig A, Liu S, et al. Transcriptomic and cellular decoding of regional brain vulnerability to neurogenetic disorders. Nat Commun, 2020,
11: 5936
33 Li J, Seidlitz J, Suckling J, et al. Cortical structural differences in major depressive disorder correlate with cell type-specific transcriptional
signatures. Nat Commun, 2021, 12: 1649
34 Mechelli A, Friston K J, Frackowiak R S, et al. Structural covariance in the human cortex. J Neurosci, 2005, 25: 8303–8310
35 Khundrakpam B S, Reid A, Brauer J, et al. Developmental changes in organization of structural brain networks. Cerebral Cortex, 2013, 23: 2072–
2085
36 Tzourio-Mazoyer N, Landeau B, Papathanassiou D, et al. Automated anatomical labeling of activations in SPM using a macroscopic anatomical
parcellation of the MNI MRI single-subject brain. NeuroImage, 2002, 15: 273–289
37 Desikan R S, Ségonne F, Fischl B, et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based
regions of interest. NeuroImage, 2006, 31: 968–980
38 Collins D L, Holmes C J, Peters T M, et al. Automatic 3-D model-based neuroanatomical segmentation. Hum Brain Mapp, 1995, 3: 190–208
39 Zheng W H. Research on methodologies of structural MRI-based connectivity extraction and their applications (in Chinese). Doctor Dissertation.
Lanzhou: Lanzhou University, 2018 [郑炜豪.基于结构磁共振影像的脑连接量化方法及应用研究.博士学位论文.兰州:兰州大学, 2018]
40 Canales-Rodríguez E J, Radua J, Pomarol-Clotet E, et al. Statistical analysis of brain tissue images in the wavelet domain: Wavelet-based
morphometry. NeuroImage, 2013, 72: 214–226
41 Kullback S, Leibler R A. On information and sufficiency. Ann Math Statist, 1951, 22: 79–86
42 Lin J. Divergence measures based on the shannon entropy. IEEE Trans Inform Theor, 1991, 37: 145–151
43 Li Y Z, Wang N K, Wang H, et al. Surface-based single-subject morphological brain networks: Effects of morphological index, brain parcellation
and similarity measure, sample size-varying stability and test-retest reliability. NeuroImage, 2021, 235: 118018
44 Khundrakpam B S, Lewis J D, Jeon S, et al. Exploring individual brain variability during development based on patterns of maturational coupling
of cortical thickness: A longitudinal MRI study. Cerebral Cortex, 2019, 29: 178–188
45 Li W, Yang C L, Shi F, et al. Construction of individual morphological brain networks with multiple morphometric features. Front Neuroanat,
2017, 11: 34
46 Yu K X, Wang X T, Li Q L, et al. Individual morphological brain network construction based on multivariate Euclidean distances between brain
regions. Front Hum Neurosci, 2018, 12: 204
47 King D J, Wood A G. Clinically feasible brain morphometric similarity network construction approaches with restricted magnetic resonance
imaging acquisitions. Network Neurosci, 2020, 4: 274–291
48 Yang S, Wagstyl K, Meng Y, et al. Cortical patterning of morphometric similarity gradient reveals diverged hierarchical organization in sensory-
motor cortices. Cell Rep, 2021, 36: 109582
49 Zhao K, Zheng Q, Che T T, et al. Regional radiomics similarity networks (R2SNs) in the human brain: Reproducibility, small-world properties and
a biological basis. Netw Neurosci, 2021, 5: 783–797
50 Fenchel D, Dimitrova R, Seidlitz J, et al. Development of microstructural and morphological cortical profiles in the neonatal brain. Cerebral
Cortex, 2020, 30: 5767–5779
51 Galdi P, Blesa M, Stoye D Q, et al. Neonatal morphometric similarity mapping for predicting brain age and characterizing neuroanatomic variation
associated with preterm birth. NeuroImage-Clin, 2020, 25: 102195
12 https://engine.scichina.com/doi/10.1360/TB-2022-0621
52 Craddock R C, Jbabdi S, Yan C G, et al. Imaging human connectomes at the macroscale. Nat Methods, 2013, 10: 524–539
53 Maier-Hein K H, Neher P F, Houde J C, et al. The challenge of mapping the human connectome based on diffusion tractography. Nat Commun,
2017, 8: 1349
54 Liang X, Wang J Y, He Y. Human connectome: Structural and functional brain networks (in Chinese). Chin Sci Bull, 2010, 55: 1565–1583 [梁夏,
王金辉,贺永.人脑连接组研究:脑结构网络和脑功能网络.科学通报, 2010, 55: 1565–1583]
55 Rubinov M, Sporns O. Complex network measures of brain connectivity: Uses and interpretations. NeuroImage, 2010, 52: 1059–1069
56 van den Heuvel M P, Sporns O. Rich-club organization of the human connectome. J Neurosci, 2011, 31: 15775–15786
57 Newman M E J, Girvan M. Finding and evaluating community structure in networks. Phys Rev E, 2004, 69: 026113
58 Stiles J, Jernigan T L. The basics of brain development. Neuropsychol Rev, 2010, 20: 327–348
59 Collin G, van den Heuvel M P. The ontogeny of the human connectome. Neuroscientist, 2013, 19: 616–628
60 Woodburn M, Bricken C L, Wu Z, et al. The maturation and cognitive relevance of structural brain network organization from early infancy to
childhood. NeuroImage, 2021, 238: 118232
61 Fenchel D, Dimitrova R, Robinson E C, et al. Neonatal multi-modal cortical profiles predict 18-month developmental outcomes. Dev Cogn
Neurosci, 2022, 54: 101103
62 Xu F F, Liu M T, Kim S Y, et al. Morphological development trajectory and structural covariance network of the human fetal cortical plate during
the early second trimester. Cerebral Cortex, 2021, 31: 4794–4807
63 Garcia K E, Robinson E C, Alexopoulos D, et al. Dynamic patterns of cortical expansion during folding of the preterm human brain. Proc Natl
Acad Sci USA, 2018, 115: 3156–3161
64 Gholipour A, Estroff J A, Warfield S K. Robust super-resolution volume reconstruction from slice acquisitions: Application to fetal brain MRI.
IEEE Trans Med Imag, 2010, 29: 1739–1758
65 Kainz B, Steinberger M, Wein W, et al. Fast volume reconstruction from motion corrupted stacks of 2D slices. IEEE Trans Med Imag, 2015, 34:
1901–1913
66 Kuklisova-Murgasova M, Quaghebeur G, Rutherford M A, et al. Reconstruction of fetal brain MRI with intensity matching and complete outlier
removal. Med Image Anal, 2012, 16: 1550–1564
67 Makropoulos A, Gousias I S, Ledig C, et al. Automatic whole brain MRI segmentation of the developing neonatal brain. IEEE Trans Med Imag,
2014, 33: 1818–1831
68 Basser P J, Mattiello J, LeBihan D. MR diffusion tensor spectroscopy and imaging. Biophysl J, 1994, 66: 259–267
69 Jensen J H, Helpern J A, Ramani A, et al. Diffusional kurtosis imaging: The quantification of non-gaussian water diffusion by means of magnetic
resonance imaging. Magn Reson Med, 2005, 53: 1432–1440
70 Zhang H, Schneider T, Wheeler-Kingshott C A, et al. NODDI: Practical in vivo neurite orientation dispersion and density imaging of the human
brain. NeuroImage, 2012, 61: 1000–1016
71 Whitaker K J, Vértes P E, Romero-Garcia R, et al. Adolescence is associated with genomically patterned consolidation of the hubs of the human
brain connectome. Proc Natl Acad Sci USA, 2016, 113: 9105–9110
72 Yeo B T T, Krienen F M, Sepulcre J, et al. The organization of the human cerebral cortex estimated by intrinsic functional connectivity. J
NeuroPhysiol, 2011, 106: 1125–1165
73 Váša F, Seidlitz J, Romero-Garcia R, et al. Adolescent tuning of association cortex in human structural brain networks. Cerebral Cortex, 2018, 28:
281–294
74 Vijayakumar N, Ball G, Seal M L, et al. The development of structural covariance networks during the transition from childhood to adolescence.
Sci Rep, 2021, 11: 1–2
75 Glasser M F, Coalson T S, Robinson E C, et al. A multi-modal parcellation of human cerebral cortex. Nature, 2016, 536: 171–178
76 van Wijk B C M, Stam C J, Daffertshofer A. Comparing brain networks of different size and connectivity density using graph theory. PLoS One,
2010, 5: e13701
77 Sporns O. The human connectome: A complex network. Ann New York Acad Sci, 2011, 1224: 109–125
78 Shi F, Yap P T, Wu G R, et al. Infant brain atlases from neonates to 1- and 2-year-olds. PLoS One, 2011, 6: e18746
79 Gholipour A, Rollins C K, Velasco-Annis C, et al. A normative spatiotemporal MRI atlas of the fetal brain for automatic segmentation and analysis
of early brain growth. Sci Rep, 2017, 7: 1–3
80 Liang X, Wang J H, Yan C G, et al. Different correlation metrics reveal different topological patterns in the human brain functional networks.
NeuroImage, 2009, 47: S170
81 Zhang Z Q, Liao W, Xu Q, et al. Hippocampus-associated causal network of structural covariance measuring structural damage progression in
temporal lobe epilepsy. Hum Brain Mapp, 2017, 38: 753–766
82 Xu Y, Xu Q, Zhang Q R, et al. Influence of epileptogenic region on brain structural changes in rolandic epilepsy. Brain Imag Behav, 2022, 16: 424–
13
https://engine.scichina.com/doi/10.1360/TB-2022-0621
434
83 Honey C J, Sporns O, Cammoun L, et al. Predicting human resting-state functional connectivity from structural connectivity. Proc Natl Acad Sci
USA, 2009, 106: 2035–2040
84 Paquola C, Vos De Wael R, Wagstyl K, et al. Microstructural and functional gradients are increasingly dissociated in transmodal cortices. PLoS
Biol, 2019, 17: e3000284
85 Margulies D S, Ghosh S S, Goulas A, et al. Situating the default-mode network along a principal gradient of macroscale cortical organization. Proc
Natl Acad Sci USA, 2016, 113: 12574–12579
86 Valk S L, Xu T, Margulies D S, et al. Shaping brain structure: Genetic and phylogenetic axes of macroscale organization of cortical thickness. Sci
Adv, 2020, 6: eabb3417
87 Xia Y M, Xia M R, Liu J, et al. Development of functional connectome gradients during childhood and adolescence. Sci Bull, 2022, 67: 1049–1061
88 Dong H M, Margulies D S, Zuo X N, et al. Shifting gradients of macroscale cortical organization mark the transition from childhood to
adolescence. Proc Natl Acad Sci USA, 2021, 118: e2024448118
89 Romero-Garcia R, Whitaker K J, Váša F, et al. Structural covariance networks are coupled to expression of genes enriched in supragranular layers
of the human cortex. NeuroImage, 2018, 171: 256–267
90 Zhang H X, Lai C, Liu R B, et al. Age-specific optimization of T1-weighted brain MRI throughout infancy. NeuroImage, 2019, 199: 387–395
91 Makropoulos A, Counsell S J, Rueckert D. A review on automatic fetal and neonatal brain MRI segmentation. NeuroImage, 2018, 170: 231–248
14 https://engine.scichina.com/doi/10.1360/TB-2022-0621
Summary for 人脑形态网络及其在脑发育研究中的应用
Brain morphological network and its applications in human
brain development
Ruoke Zhao1, Zhiyong Zhao1, Jinhui Wang2& Dan Wu1*
1College of Biomedical Engineering and Instrument Science, Zhejiang University, Hangzhou 310027, China;
2Institute for Brain Research and Rehabilitation, South China Normal University, Guangzhou 510631, China
* Corresponding author, E-mail: danwu.bme@zju.edu.cn
The human brain is an extremely complex system, where trillions of neurons are connected through synapses, forming a
massive micro-scale anatomical network which is considered to be the physiological basis for information processing and
cognition. How to depict the connection pattern of the brain has always been a hot topic within the neuroscience
community. With the development of non-invasive neuroimaging technology, investigations of the brain network have
entered a new era. Brain networks based on magnetic resonance imaging (MRI) provide new insights into the complex
anatomical organization and functional mechanism of the human brain.
Brain networks can be divided into morphological networks, white matter networks, and functional networks. Brain
morphological networks usually refer to the morphological connection networks generated from structural MRI, and have
received wide attention because of its advantages of easy image acquisition, stable image quality, and simplicity in analysis.
Brain morphological networks can be further divided into across-subjects covariance networks and individual-level
networks. Across-subjects covariance networks are constructed by calculating the statistical correlation of morphological
features between brain regions from a group of participants, which only reflect the group characteristics of participants but
ignore the individual variations. Individual-level brain morphological networks are built using morphological information
from the individual brain to describe the cooperative changes between brain regions, which can preserve the individual
difference. There are two types of methods to build individual morphological networks, namely, single feature based and
multiple features based, which will be described in detail in this review.
Graph theory model in combination with brain morphological network enables us to characterize the developmental
trajectories of the topology of cortical morphology, from prenatal, early postnatal, childhood to adolescence. Based on the
recent studies of brain morphological networks on brain development, we found that brain morphological networks are still
in their primitive state during the early second trimester, which become more integrated towards birth and already exhibit
the characteristic small-world topology and nonrandom modular organization at birth. Subsequently, brain morphological
networks become more separated from birth to early childhood and gradually integrate from late childhood to adolescence.
In addition, many studies demonstrated distinct developmental trajectories in different cortical divisions. Specifically, the
primary networks are the first to reach maturity, while the higher-order networks continue to develop for a long time. These
findings provide important evidence for understanding the reorganization of the cerebral cortex and the emergence and
development of brain cognitive function.
Although brain morphological networks have undergone ten years of development, problems and challenges remain in this
field. Similar to other forms of connectivity analysis, brain morphological networks analysis struggles with the issues of edge
and node definition, and further methodological development is needed in the fields of graph theory and statistical comparison
between networks. More importantly, despite some studies that attempt to relate brain morphological networks to tractography-
based brain networks, functional networks, and gene-specific expression, we are still awaiting a clear interpretation of the
biological significance of brain morphological networks. In addition, how to acquire high-quality MRI data for perinatal brains
is a major challenge to complete the development trajectories of developing brain morphological networks.
magnetic resonance imaging, brain morphological network, topological organization, brain development,
segregation and integration
doi: 10.1360/TB-2022-0621
15
https://engine.scichina.com/doi/10.1360/TB-2022-0621
Article
Full-text available
Aims The human brain is an extremely complex system in which neurons, clusters of neurons, or regions are connected to form a complex network. With the development of neuroimaging techniques, magnetic resonance imaging (MRI)‐based brain networks play a key role in our understanding of the intricate architecture of human brain. Among them, the structural MRI‐based brain morphological network approach has attracted increasing attention due to the advantages in data acquisition, image quality, and in revealing the structural organizing principles intrinsic to the brain. This review is to summarize the methodology and related applications of individual‐level morphological networks. Background There have been a growing number of studies related to brain morphological similarity networks. Conventional morphological networks are intersubject covariance networks constructed using a certain morphological indicator of a group of subjects; individual‐level morphological networks, on the other hand, measure the morphological similarity between brain regions for individual brains and can reflect the morphological information of single subjects. In recent years, individual morphological networks have demonstrated significant worth in exploring the topological changes of the human brain under both normal and disease conditions. Such studies provided novel perspectives for understanding human brain development and exploring the pathological mechanisms of neuropsychiatric disorders. Conclusion This paper mainly focuses on the studies of brain morphological networks at the individual level, introduces several ways for network construction, reviews representative work in this field, and finally points out current problems and future directions.
Article
Full-text available
Developmental delays in infanthood often persist, turning into life-long difficulties, and coming at great cost for the individual and community. By examining the developing brain and its relation to developmental outcomes we can start to elucidate how the emergence of brain circuits is manifested in variability of infant motor, cognitive and behavioural capacities. In this study, we examined if cortical structural covariance at birth, indexing coordinated development, is related to later infant behaviour. We included 193 healthy term-born infants from the Developing Human Connectome Project (dHCP). An individual cortical connectivity matrix derived from morphological and microstructural features was computed for each subject (morphometric similarity networks, MSNs) and was used as input for the prediction of behavioural scores at 18 months using Connectome-Based Predictive Modeling (CPM). Neonatal MSNs successfully predicted social-emotional performance. Predictive edges were distributed between and within known functional cortical divisions with a specific important role for primary and posterior cortical regions. These results reveal that multi-modal neonatal cortical profiles showing coordinated maturation are related to developmental outcomes and that network organization at birth provides an early infrastructure for future functional skills.
Article
Full-text available
Connectome mapping studies have documented a principal primary-to-transmodal gradient in the adult brain network, capturing a functional spectrum that ranges from perception and action to abstract cognition. However, how this gradient pattern develops and whether its development is linked to cognitive growth, topological reorganization, and gene expression profiles remain largely unknown. Using longitudinal resting-state functional magnetic resonance imaging data from 305 children (aged 6–14 years), we describe substantial changes in the primary-to-transmodal gradient between childhood and adolescence, including emergence as the principal gradient, expansion of global topography, and focal tuning in primary and default-mode regions. These gradient changes are mediated by developmental changes in network integration and segregation, and are associated with abstract processing functions such as working memory and expression levels of calcium ion regulated exocytosis and synaptic transmission-related genes. Our findings have implications for understanding connectome maturation principles in normal development and developmental disorders.
Article
Full-text available
The topological organization of the cerebral cortex provides hierarchical axes, namely gradients, which reveal systematic variations of brain structure and function. However, the hierarchical organization of macroscopic brain morphology and how it constrains cortical function along the organizing axes remains unclear. We map the gradient of cortical morphometric similarity (MS) connectome, combining multiple features conceptualized as a “fingerprint” of an individual’s brain. The principal MS gradient is anchored by motor and sensory cortices at two extreme ends, which are reliable and reproducible. Notably, divergences between motor and sensory hierarchies are consistent with the laminar histological thickness gradient but contrary to the canonical functional connectivity (FC) gradient. Moreover, the MS dissociates with FC gradients in the higher-order association cortices. The MS gradient recapitulates fundamental properties of cortical organization, from gene expression and cyto- and myeloarchitecture to evolutionary expansion. Collectively, our findings provide a heuristic hierarchical organization of cortical morphological neuromarkers.
Article
Full-text available
To investigate the influence of epileptogenic cortex (Rolandic areas) with executive functions in Rolandic epilepsy using structural covariance analysis of structural magnetic resonance imaging (MRI). Structural MRI data of drug-naive patients with Rolandic epilepsy (n = 70) and typically developing children as healthy controls (n = 83) were analyzed using voxel-based morphometry. Gray matter volumes in the patients were compared with those of healthy controls, and were further correlated with epilepsy duration and cognitive score of executive function, respectively. By applying Granger causal analysis to the sequenced morphometric data according to disease progression information, causal network of structural covariance was constructed to assess the causal influence of structural changes from Rolandic cortices to the regions engaging executive function in the patients. Compared with healthy controls, epilepsy patients showed increased gray matter volume in the Rolandic regions, and also the regions engaging in executive function. Covariance network analyses showed that along with disease progression, the Rolandic regions imposed positive causal influence on the regions engaging in executive function. In the patients with Rolandic epilepsy, epileptogenic regions have causal influence on the structural changes in the regions of executive function, implicating damaging effects of Rolandic epilepsy on human brain.
Article
Full-text available
Significance Here, we describe age-dependent shifts in the macroscale organization of cortex in childhood and adolescence. The characterization of functional connectivity patterns in children revealed an overarching organizational framework anchored within the unimodal cortex, between somatosensory/motor and visual regions. Conversely, in adolescents, we observed a transition into an adult-like gradient, situating the default network at the opposite end of a spectrum from primary somatosensory/motor regions. This spatial framework emerged gradually with age, reaching a sharp inflection point at the transition from childhood to adolescence. These data reveal a developmental change from a functional motif first dominated by the distinction between sensory and motor systems and then balanced through interactions with later-maturing aspects of association cortex that support more abstract cognitive functions.
Article
Full-text available
A structural covariance network (SCN) has been used successfully in structural magnetic resonance imaging (sMRI) studies. However, most SCNs have been constructed by a unitary marker that is insensitive for discriminating different disease phases. The aim of this study was to devise a novel regional radiomics similarity network (R2SN) that could provide more comprehensive information in morphological network analysis. R2SNs were constructed by computing the Pearson correlations between the radiomics features extracted from any pair of regions for each subject. We further assessed the small-world properties of R2SNs, and we evaluated the reproducibility in different datasets and through test-retest analysis. The relationships between the R2SNs and general intelligence/interregional coexpression of enriched genes were also explored. R2SNs could be replicated in different datasets, regardless of the use of different feature subsets. R2SNs showed high reproducibility in the test-retest analysis (intraclass correlation coefficient >0.7). In addition, the small-word property (σ>2) and the high correlation between gene expression (R=0.29, P<0.001) and general intelligence were determined for R2SNs. Furthermore, the results have also been repeated in the Brainnetome atlas. R2SNs provide a novel, reliable and biologically plausible method to understand human morphological covariance based on sMRI.
Article
Full-text available
Structural covariance conceptualizes how morphologic properties of brain regions are related to one another (across individuals). It can provide unique information to cortical structure (e.g., thickness) about the development of functionally meaningful networks. The current study investigated how structural covariance networks develop during the transition from childhood to adolescence, a period characterized by marked structural re-organization. Participants (N = 192; scans = 366) completed MRI assessments between 8.5 and 14.5 years of age. A sliding window approach was used to create “age-bins”, and structural covariance networks (based on cortical thickness) were created for each bin. Next, generalized additive models were used to characterize trajectories of age-related changes in network properties. Results revealed nonlinear trajectories with “peaks” in mean correlation and global density that are suggestive of a period of convergence in anatomical properties across the cortex during early adolescence, prior to regional specialization. “Hub” regions in sensorimotor cortices were present by late childhood, but the extent and strength of association cortices as “hubs” increased into mid-adolescence. Moreover, these regional changes were found to be related to rates of thinning across the cortex. In the context of neurocognitive networks, the frontoparietal, default mode, and attention systems exhibited age-related increases in within-network and between-network covariance. These regional and modular developmental patterns are consistent with continued refinement of socioemotional and other complex executive functions that are supported by higher-order cognitive networks during early adolescence.
Article
Full-text available
Morphological brain networks, in particular those at the individual level, have become an important approach for studying the human brain connectome; however, relevant methodology is far from being well-established in their formation, description and reproducibility. Here, we extended our previous study by constructing and characterizing single-subject morphological similarity networks from brain volume to surface space and systematically evaluated their reproducibility with respect to effects of different choices of morphological index, brain parcellation atlas and similarity measure, sample size-varying stability and test-retest reliability. Using the Human Connectome Project dataset, we found that surface-based single-subject morphological similarity networks shared common small-world organization, high parallel efficiency, modular architecture and bilaterally distributed hubs regardless of different analytical strategies. Nevertheless, quantitative values of all interregional similarities, global network measures and nodal centralities were significantly affected by choices of morphological index, brain parcellation atlas and similarity measure. Moreover, the morphological similarity networks varied along with the number of participants and approached stability until the sample size exceeded ∼70. Using an independent test-retest dataset, we found fair to good, even excellent, reliability for most interregional similarities and network measures, which were also modulated by different analytical strategies, in particular choices of morphological index. Specifically, fractal dimension and sulcal depth outperformed gyrification index and cortical thickness, higher-resolution atlases outperformed lower-resolution atlases, and Jensen-Shannon divergence-based similarity outperformed Kullback-Leibler divergence-based similarity. Altogether, our findings propose surface-based single-subject morphological similarity networks as a reliable method to characterize the human brain connectome and provide methodological recommendations and guidance for future research.
Article
The interactions of brain regions with other regions at the network level likely provide the infrastructure necessary for cognitive processes to develop. Specifically, it has been theorized that in infancy brain networks become more modular, or segregated, to support early cognitive specialization, before integration across networks increases to support the emergence of higher-order cognition. The present study examined the maturation of structural covariance networks (SCNs) derived from longitudinal cortical thickness data collected between infancy and childhood (0-6 years). We assessed modularity as a measure of network segregation and global efficiency as a measure of network integration. At the group level, we observed trajectories of increasing modularity and decreasing global efficiency between early infancy and six years. We further examined subject-based maturational coupling networks (sbMCNs) in a subset of this cohort with cognitive outcome data at 8-10 years, which allowed us to relate the network organization of longitudinal cortical thickness maturation to cognitive outcomes in middle childhood. We found that lower global efficiency of sbMCNs throughout early development (across the first year) related to greater motor learning at 8-10 years. Together, these results provide novel evidence characterizing the maturation of brain network segregation and integration across the first six years of life, and suggest that specific trajectories of brain network maturation contribute to later cognitive outcomes.
Article
During the early second trimester, the cortical plate, or “the developing cortex”, undergoes immensely complex and rapid development to complete its major complement of neurons. However, morphological development of the cortical plate and the precise patterning of brain structural covariance networks during this period remain unexplored. In this study, we used 7.0 T high-resolution magnetic resonance images of brain specimens ranging from 14 to 22 gestational weeks to manually segment the cortical plate. Thickness, area expansion, and curvature (i.e., folding) across the cortical plate regions were computed, and correlations of thickness values among different cortical plate regions were measured to analyze fetal cortico-cortical structural covariance throughout development of the early second trimester. The cortical plate displayed significant increases in thickness and expansions in area throughout all regions but changes of curvature in only certain major sulci. The topological architecture and network properties of fetal brain covariance presented immature and inefficient organizations with low degree of integration and high degree of segregation. Altogether, our results provide novel insight on the developmental patterning of cortical plate thickness and the developmental origin of brain network architecture throughout the early second trimester.