| 徐志豪,孙莉敏,王鹤玮,沈逸凡,杨以涵,郭耀今,卢 杉,刘志超,黎启光,余秋蓉,李哲宇,尹大志,范明霞.基于间接断连评估结构损伤对脑卒中后上肢运动功能及康复潜力的影响[J].中国康复医学杂志,2026,(8):1200~1209 |
| 基于间接断连评估结构损伤对脑卒中后上肢运动功能及康复潜力的影响 点此下载全文 |
| 徐志豪 孙莉敏 王鹤玮 沈逸凡 杨以涵 郭耀今 卢 杉 刘志超 黎启光 余秋蓉 李哲宇 尹大志 范明霞 |
| 华东师范大学物理学院上海市磁共振重点实验室，上海市，200062 |
| 基金项目：国家自然科学基金面上项目（81974356）；上海市科委自然基金面上项目（23ZR1408500）；上海市卫生健康系统重点扶持学科建设项目（2023ZDFC0304）；福建省科技厅自然科学基金项目（2025J01724）；福建省科技创新联合资金项目（2025Y9138）；国家自然科学基金青年项目（82102665）；上海市科学技术委员会“扬帆计划”项目（21YF1404600） |
| DOI：10.3969/j.issn.1001-1242.2026.08.003 |
| 摘要点击次数: 84 |
| 全文下载次数: 21 |
| 摘要： |
| 摘要
目的：探讨脑卒中患者全脑结构断连特征，并验证其在预测运动功能受损严重程度及康复潜力方面是否优于局部病灶体积。
方法：纳入60例首次发病的单侧皮层下脑卒中患者，利用间接断连分析的NeMo（Network Modification）工具包， 将其T1磁共振成像上的病灶掩模叠加至人类连接组计划（human connectome project， HCP）健康人群脑连接图谱，计算全脑结构断连连通性变化（change in connectivity,ChaCo）分数。经主成分分析（principal component analysis， PCA）后，结合岭回归构建ChaCo模型与病灶体积模型，对比两者对康复干预前后上肢运动功能评分（Fugl-Meyer assessment of upper extremity，FMA-UE）的解释力，并提取受累脑区贡献权重。
结果：全脑断连呈显著偏侧性，病灶同侧皮层下区域（基底核、丘脑等）损伤较皮层更严重。ChaCo模型可有效预测基线期（R2=0.394，r=0.660，P=0.014）及干预4周后（R2=0.412，r=0.667，P=0.016）的FMA-UE评分，且效能显著优于病灶体积模型，干预前两者的解释力差值ΔR2为0.140（P=0.016），干预后差值提升至0.208（P=0.019）。康复干预后，脑区权重从皮层下驱动转向额叶、顶叶等皮层主导，感觉运动相关皮层脑区权重上升与运动功能改善一致。
结论：脑结构断连分析的脑区权重定向变化可作为脑卒中康复疗效的重要神经影像学征象。全脑结构断连特征在运动功能预测上优于传统病灶体积指标，可为康复疗效评估与方案优化提供参考依据。 |
| 关键词：脑卒中 结构断连 连通性变化分数 运动功能受损 |
| Evaluating the impact of structural damage on post-stroke upper extremity motor function and rehabilitation potential via indirect disconnection assessment Download Fulltext |
|
| East China Normal University， School of Physics， Shanghai， 200062 |
| Fund Project: |
| Abstract: |
| Abstract
Objective： To investigate the characteristics of whole-brain structural disconnection in stroke patients and to verify whether disconnectome-based metrics provide better predictive performance than focal lesion volume for upper extremity motor impairment and rehabilitation potential.
Method： Sixty patients with first-ever unilateral subcortical stroke were enrolled. Using the network modification (NeMo) tool for indirect disconnection analysis, lesion masks derived from T1-weighted magnetic resonance imaging (MRI) were superimposed onto the healthy structural connectome template from the Human Connectome Project HCP) to calculate whole-brain structural disconnection, quantified as change in connectivity (ChaCo) scores. Subsequently， Principal Component Analysis （PCA） combined with ridge regression was used to construct both the ChaComodel and the lesion volume model. The explanatory power of the two models for the Fugl-Meyer assessment of upper extremity （FMA-UE） scores before and after rehabilitation intervention was compared， and the contribution weights of the affected brain regions were extracted.
Result： Whole-brain structural disconnection exhibited significant laterality， with more severe damage in ipsilesional subcortical regions （e.g.， basal ganglia， thalamus） compared to the cortex. The ChaComodel efficiently predicted FMA-UE scores at baseline （R2=0.394，r=0.660，P=0.014） and 4 weeks post-intervention （R2=0.412， r=0.667，P=0.016）. Its predictive performance was significantly superior to that of the lesion-volume model. The difference in explained variance（ΔR2） between the two models was 0.140 （P=0.016） pre-intervention， increasing to 0.208 （P=0.019） post-intervention. Following rehabilitation intervention， the regional predictive weights shifted from being subcortically-driven to cortically-dominated （e.g.， frontal and parietal lobes）， and the increased weights in sensorimotor-related cortical areas were consistent with motor functional improvement.
Conclusion： The directional shifts in regional weights derived from brain structural disconnection analysis can serve as important neuroimaging biomarkers for stroke rehabilitation efficacy. Whole-brain structural disconnectivity features outperform traditional lesion volume metrics in predicting motor function， providing a robust reference for evaluating rehabilitation outcomes and optimizing treatment protocols. |
| Keywords:stroke structural disconnection change in connectivity score motor impairment |
|
| 查看全文 查看/发表评论 |
