Page 76 - 南京医科大学自然版

南京医科大学学报（自然科学版） 第45卷第8期
·1132 · Journal of Nanjing Medical University（Natural Sciences） 2025年8月

·临床研究·

F⁃FDG⁃PET/MR 探究早期帕金森病脑代谢网络和功能网络改
18
变特征

张逸悦 ，李孝媛 ，高 擎 ，王 峰 ，王柏林 ，李如帅 ，张朋俊 1
1
1
2
1
1
1*
南京医科大学附属南京医院（南京市第一医院）核医学科，神经内科，江苏 南京 210006
1 2
［摘 要］ 目的：用 F⁃FDG⁃PET/MR 和詹森⁃香农相似性估计（Jensen⁃Shannon similarity estimation，JSSE）方法构建帕金森病
18
（Parkinson’s disease，PD）个体脑代谢网络，探究网络拓扑属性的变化及其与大脑功能网络之间的相关性。方法：回顾性分析2022年
8月—2024年5月在南京医科大学附属南京医院行 F⁃FDG⁃PET/MR显像且经临床确诊的PD患者28例，男16例，女12例；同时
18
纳入与PD患者年龄与性别匹配的正常对照组（normal control，NC）20例，男10例，女10例。采用JSSE方法构建个体脑代谢网
络，Matlab2013b利用Gretna分析网络拓扑属性和脑区之间代谢连接的变化。使用两独立样本t检验及Mann⁃Whitney U检验分
析数据，通过Mantel检验分析PD组以及NC组代谢网络与功能网络矩阵之间的相关性。结果：与NC组相比，PD组在全局属性
方面表现出较高的同配性（t=-2.06，P=0.045）；在节点属性方面，右侧梭状回节点度中心性显著降低（t=3.32，P < 0.001）；共有
15个关键代谢连接参数降低（P < 0.001），其中左侧尾状核和左侧颞下回之间代谢连接减少最明显。PD代谢网络和功能网络矩阵
之间存在显著的正相关（r=0.242，P < 0.05），相关性大于NC组（r=0.197，P < 0.05）。结论：PET/MR和 JSSE能早期识别PD脑拓扑
属性和代谢连接异常，F⁃FDG⁃PET/MR显像有助于PD早期诊断，从脑代谢与功能相关角度探究PD患者脑部内在重塑机制变化。
18
［关键词］ PET/MR；詹森⁃香农相似性估计；代谢脑网络；相关系数；帕金森病
［中图分类号］ R742.5；R817.4 ［文献标志码］ A ［文章编号］ 1007⁃4368（2025）08⁃1132⁃08
doi：10.7655/NYDXBNSN250119

F⁃FDG⁃PET/MR unveils altered features of brain metabolic and functional networks in
18
early Parkinson’s disease
1
1
1
1*
1
2
ZHANG Yiyue ，LI Xiaoyuan ，GAO Qing ，WANG Feng ，WANG Bailin ，LI Rushuai ，ZHANG Pengjun 1
1 Department of Nuclear Medicine，Department of Neurology，the Affiliated Nanjing Hospital of Nanjing Medical
2
University（Nanjing First Hospital），Nanjing 210006，China
［Abstract］ Objective：This study aims to establish individual cerebral metabolic networks in Parkinson’s disease（PD）using
18
integrated F ⁃ FDG ⁃ PET/MR imaging and Jensen ⁃ Shannon similarity estimation（JSSE）and investigate topological alterations in
metabolic networks and their associations with functional networks. Methods：Twenty⁃eight patients with clinically confirmed PD（16
18
males and 12 females）who underwent F⁃FDG⁃PET/MR imaging at the Affiliated Nanjing Hospital of Nanjing Medical University from
August 2022 to May 2024 were retrospectively analyzed，and 20 age⁃ and sex⁃matched normal controls（NC）were included（10 males
and 10 females）. Individual brain metabolic networks were constructed using the JSSE method，and Matlab2013b was used to analyze
changes between network topological properties and metabolic connectivity in brain regions using Gretna. Data were analyzed using two
independent samples t⁃test and Mann⁃Whitney U⁃test，and the Mantel test analyzed correlations between metabolic and functional
network matrices in the PD group and the NC group. Results：Compared with the NC group，the PD group showed higher assortativity
in terms of global attributes（t=-2.06，P=0.045）；in terms of nodal attributes，the centrality of nodal degree in the right fusiform gyrus
was significantly reduced（t=3.32，P < 0.001）；15 key metabolic connectivity parameters were reduced（P < 0.001），with the reduction
in metabolic connectivity between the left caudate nucleus and the left inferotemporal gyrus being the most pronounced. There was a
significant positive correlation between the PD metabolic and functional network matrices（r=0.242，P < 0.05），and the correlation was

［基金项目］ 国家重点研发计划（2022YFC2406900）；江苏省社会发展基金（BE2017612）；江苏省医学重点学科（JSDW202247）
通信作者（Corresponding author），E⁃mail：fengwangcn@hotmail.com（ORCID：0000⁃0003⁃1089⁃3095）
∗