Kuang Cuili, Zha Yunfei, Liu Changsheng, Chen Jun
Department of Radiology, Renmin Hospital of Wuhan University, Wuhan, China.
Front Hum Neurosci. 2020 Sep 4;14:364. doi: 10.3389/fnhum.2020.00364. eCollection 2020.
Brain structural alterations play an important role in patients with cervical spondylotic myelopathy (CSM). However, while there have been studies on regional brain structural alterations, only few studies have focused on the topological organization of the brain structural covariance network. This work aimed to describe the structural covariance network architecture alterations that are possibly linked to cortex reorganization in patients with CSM.
High-resolution anatomical images of 31 CSM patients and 31 healthy controls (HCs) were included in the study. The images were acquired using a sagittal three-dimensional T1-weighted BRAVO sequence. Firstly, the gray matter volume of 90 brain regions of automated anatomical labeling atlas were computed using a VBM toolbox based on the DARTEL algorithm. Then, the brain structural covariance network was constructed by thresholding the gray matter volume correlation matrices. Subsequently, the network measures and nodal property were calculated based on graph theory. Finally, the differences in the network metrics and nodal property between groups were compared using a non-parametric test.
Patients with CSM showed larger global efficiency and smaller local efficiency, clustering coefficient, characteristic path length, and sigma values than HCs. Patients with CSM had greater betweenness in the left superior parietal gyrus (SPG.L) and the left supplementary motor area (SMA.L) than HCs. Besides, patients with CSM had smaller betweenness in right middle occipital gyrus. The brain structural covariance networks of CSM patients exhibited equal resilience to random failure as those of HCs. However, the maximum relative size of giant connected components was approximately 10% larger in HCs than in CSM patients, upon removal of 44 nodes in targeted attack.
These observed alternations in global network measures in CSM patients reflect that the brain structural covariance network in CSM exhibits the less optimal small-world model compared to that in HCs. Increased betweenness in SPG.L and SMA.L seems to be related to cortex reorganization to recover multiple sensory functions after spinal cord injury in CSM patients. The network resilience of patients with CSM exhibiting a relative mild vulnerability, compared to HCs, is probably attributable to the balance and interplay between cortex reorganization and ongoing degeneration.
脑结构改变在脊髓型颈椎病（CSM）患者中起重要作用。然而，虽然已有关于脑局部结构改变的研究，但仅有少数研究关注脑结构协方差网络的拓扑组织。本研究旨在描述可能与CSM患者皮质重组相关的结构协方差网络结构改变。
本研究纳入了31例CSM患者和31例健康对照（HCs）的高分辨率解剖图像。图像采用矢状三维T1加权BRAVO序列采集。首先，使用基于DARTEL算法的VBM工具箱计算自动解剖标记图谱中90个脑区的灰质体积。然后，通过对灰质体积相关矩阵进行阈值处理构建脑结构协方差网络。随后，基于图论计算网络指标和节点属性。最后，使用非参数检验比较两组之间网络指标和节点属性的差异。
与HCs相比，CSM患者表现出更高的全局效率以及更低的局部效率、聚类系数、特征路径长度和西格玛值。与HCs相比，CSM患者在左侧顶上叶（SPG.L）和左侧辅助运动区（SMA.L）具有更高的介数中心性。此外，CSM患者在右侧枕中回的介数中心性较小。CSM患者的脑结构协方差网络与HCs的网络对随机故障具有同等的弹性。然而，在进行靶向攻击去除44个节点后，HCs中巨连通分量的最大相对大小比CSM患者大约大10%。
CSM患者中观察到的这些全局网络指标的改变反映出，与HCs相比，CSM患者的脑结构协方差网络表现出不太理想的小世界模型。SPG.L和SMA.L中介数中心性的增加似乎与CSM患者脊髓损伤后皮质重组以恢复多种感觉功能有关。与HCs相比，CSM患者的网络弹性表现出相对较轻的脆弱性，这可能归因于皮质重组与进行性退变之间的平衡和相互作用。
