使用 MRI 构建的大脑连接网络预测帕金森病萎缩进展模式| 文献速递-基于深度学习的乳房、前列腺疾病诊断系统
Title
题目
Brain Connectivity Networks Constructed Through MRI Plays a Crucial Role in Forecasting the Progression of Atrophy in Parkinson's Disease
使用 MRI 构建的大脑连接网络预测帕金森病萎缩进展模式
Background
背景
Investigating whether connectome mapping techniques between different brain regions could enable the anticipation of degenerative trajectories in patients with mildly affected Parkinson's disease has not yet seen extensive investigation.
研究全脑结构与功能连接组映射在轻度帕金森病患者萎缩程度预测中的可行性尚待深入探讨
Method
方法
A longitudinal cohort study was conducted to investigate participants with early-stage Parkinson's disease (PD) and control subjects enrolled at a single site between January 2012 and December 2023. Participants diagnosed with PD underwent three-dimensional T1-weighted brain magnetic resonance imaging (MRI), while measurements were taken to assess regional gray matter (GM) atrophy from baseline through three consecutive annual assessments. Structural and functional brain connectivity networks were established in healthy controls using diffusion tensor imaging (DTI) and resting-state functional MRI (rsfMRI). Disease exposure (DE) indices, representing pathological characteristics of specific brain regions, were derived from the structural or functional connectivity profiles within the healthy connectome framework, combined with measures of gray matter loss in affected regions among PD patients. Researchers examined partial correlations between DE index changes in each GM region across time points (1- or 2-year follow-up) and corresponding rates of gray matter reduction (atrophy progression) observed two to three years later. Predictive models for estimating long-term gray matter loss were developed by systematically evaluating various candidate features.
这项前瞻性研究招募了来自单一医疗中心的轻度帕金森病（PD）患者及对照组参与者。
Conclusion
结论
The anatomical structure and operational functionality of the cerebral network system are integral to the early pathologic degeneration progression during Parkinson's disease development.
大脑连接组的空间组织模式及其功能网络在帕金森病（PD）早期阶段发生萎缩的过程中发挥关键作用
Results
结果
The study included a total of 86 participants with mild Parkinson's disease (average age at MRI: 60 years ± 8 years [SD]; 48 males) and 60 healthy controls (average age at MRI: 62 years ± 9 years; 31 females). The DE indices measured at the first and second years were significantly correlated with changes in atrophy from the second to the third year (r range: 0.22–0.33; P value range: .002–.04). Multivariate models incorporating DE indices demonstrated that these markers could predict a gradual increase in gray matter loss in the right caudate nucleus as well as some frontal, parietal, and temporal brain regions over a three-year period (R² range: 0.40–0.61; all P < .001).
招募了86例轻度帕金森病患者（MRI评估时平均年龄为60岁±8 [SD]；其中48名为男性）及60例健康对照组（MRI评估时平均年龄为62岁±9岁；其中31名为女性）。临床病程进展程度在第1至第2年与第2至第3年的萎缩程度呈现显著正相关关系（r值范围为0.22至0.33；P值范围为.002至.04）。基于包含DE指数的多元回归模型能够预测右侧尾状核、部分额叶、顶叶以及颞叶脑区在第3年内灰质体积的累计萎缩量（R²值范围为 47-59% 且所有P值均小于. 5 。）
Figure
图
图1：展示研究人群纳入与排除标准的研究流程图。PD = Parkinson disease.
图1: 显示研究队列的纳入和排除过程的流程图。PD = 帕金森病。
_Figure 2: Study methods diagram. (A) Healthy controls' structural and functional brain connectome at baseline was constructed. (B) Cortical thickness and gray matter (GM) volume were measured on three-dimensional T1-weighted images in Mild Parkinson Disease (PD) participants (all visits) and healthy controls (baseline). (C) Disease exposure indexes—based on pathologies in each brain region calculated at 1- and 2-year follow-up—were defined as a function of structural and functional connectivity relative to healthy controls and the severity of atrophy in connected regions. (D) Differences in cortical thickness or GM volume at baseline between Mild PD participants and healthy controls were estimated, along with changes over 1, 2, and 3 years of follow-up in Mild PD participants. (E) Prediction models were used to evaluate temporal changes in GM atrophy across brain regions in Mild PD participants. fMRI = functional MRI.
图2展示了研究方法的流程图。（A）在基线时期构建了健康对照组的结构和功能脑连接组。（B）测量了轻度帕金森病（PD）患者（包括所有随访点）及其健康对照组（基线时期）在三维T1加权图像上的皮层厚度及灰质体积。（C）疾病暴露指数被定义为，在1年和2年随访期间计算出的每个脑区的病理学指标。这些指标基于健康对照组连接组中相关区域的结构与功能连接性，并结合PD患者该区域萎缩程度。（D）评估轻度帕金森病患者与健康对照组相比在基线时期的皮层厚度及灰质体积差异，并观察其随1年至3年随访期间的变化趋势。（E）通过预测模型评估轻度帕金森病患者的灰质体积变化及其大脑区域的时间演变特征。所使用的功能性MRI（fMRI）技术提供了关于大脑活动的空间定位信息。
图3展示了参与试验的小额帕金森病患者（PD）的灰质萎缩模式与健康对照组（HC）的对比结果。研究采用的是经过年龄校正的FreeSurfer中的广义线性模型分析，并分别在基线和1年、2年、3年随访阶段进行了观察。颜色条代表t值。L表示左侧区域，R表示右侧区域。（源自参考文献15.)
图3展示了轻度帕金森病(PD)患者灰质萎缩模式的研究。(A) 在基线阶段与健康对照组(HC)进行比较；(B) 在1年至3年的随访期间的情况。使用FreeSurfer中的广义线性模型对年龄进行了校正。颜色条显示t值。L代表左侧区域, R代表右侧区域。(图经参考文献15的许可改编.)
Figure 4: 散点图4展示了灰质退行性与结构连通疾病指数之间的偏相关关系（Partial correlations between gray matter (GM) atrophy progression and structural connectivity disease exposure（DE） indexes）通过使用数目流管线（NOS）和分数离析度（FA）在参与者患有轻度帕金森病的情况下计算得出。图中分别展示的是（A）1年时DE指数与2年时GM退行性；以及（B）1年时DE指数（紫色线段）与2年时GM退行性（橙色线段）。dots代表每个分析中所涉及的所有83个脑区。dotted线条表示这些变量之间关系的趋势。FU = follow-up.
图4: 散点图揭示了轻度帕金森病受试者灰质（GM）萎缩程度与其结构连接疾病暴露（DE）指数之间的部分关联性。DE指数是由计算每个GM区域的流线数量（NOS）和分数各向异性（FA）而得出。（A)部分图展示了1年时与2年时比较的结果及其相关性；(B)则比较了1年期至2年期及至3年期的相关变化趋势。每个分析中的点代表83个不同的脑区，在这些区域上观察到了变化情况；虚线则描绘了各变量之间关系的整体趋势。（注：FU表示随访时间单位）。
_Figure 5: The scatter plots demonstrate the partial correlation between gray matter (GM) progression of atrophy and functional connectivity disease exposure (DE) index across various GM regions in participants with mild Parkinson's disease. In both scenarios, partial correlations are examined: Part A focuses on the DE index measured at one year alongside GM atrophy assessed two years later, while Part B compares DE indices measured annually—specifically, one year in purple and two years in orange—with corresponding GM atrophy progressions. Dots symbolize all 83 brain regions analyzed in each case, with dotted lines illustrating the general trends of variable relationships.
图5: 该散点图呈现了轻度帕金森病患者群体中灰质（GM）萎缩程度与功能连接（FC）疾病暴露（DE）指数之间的部分相关关系。（A）部分展示了一年期DE指数与两年期GM萎缩间存在的部分相关关系；（B）则分别用不同颜色标注了一年内和两年内DE指数与三年内GM萎缩间的表现。每个分析中的数据点代表共计83个脑区。通过虚线表示出了各变量间大致的趋势关系。FU = 随访时间
Table
表
Table 1: Population Statistics and Cognitive Features at the Start in Healthy Controls and Mild PD Participants
表1 列出了健康对照组及轻度帕金森病（PD）受试者在入组时的人口统计数据与认知特征
Tableau 2: 轻度PD患者基线时的临床参数
表2: 轻度帕金森病（PD）参与者在基线时的临床特征
表
表3：基于1年期MRI和临床指标的轻度帕金森病（PD）受试者在2年及3年间灰质区域（GM）萎缩变化的预测模型
Table 4: Predictive Models of Gait Metrics at 3 Years Dependent on MRI Scans and Clinical Assessments among Individuals with Early Signs of Parkinson’s Disease.
表格4：根据12个月和24个月期间的MRI检查及临床评估指标研究对象在第三年的灰质（GM）萎缩变化预测模型