Zeng Peng, Zeng Bang, Luo Dan, Li Binglan, Peng Yuling, Xiang Yayun, Wang Dan, Chai Ying, Li Yongmei
Department of Radiology, the First Affiliated Hospital of Chongqing Medical University, Chongqing 400016, China.
Department of Radiology, the First Affiliated Hospital of Chongqing Medical University, Chongqing 400016, China; Department of Radiology, People's Hospital of Shapingba District, 44# Xiaolongkan New Street, Chongqing 400010, China.
Magn Reson Imaging. 2025 Oct;122:110425. doi: 10.1016/j.mri.2025.110425. Epub 2025 Jun 6.
To explore the alteration of neurovascular coupling (NVC), relationships between neuroimaging metrics with clinical assessments, and classification metrics in cerebral small vessel disease (CSVD).
Participants were grouped into healthy control, CSVD with normal cognition, and CSVD with cognition impairment according to CSVD scales and Montreal Cognitive Assessment. Cerebral blood flow (CBF) adjusted for arterial transit time and dynamic/static amplitude of low-frequency fluctuation (dALFF, ALFF) were combined to evaluate NVC and to determine intergroup differences. Partial Spearman Correlation between measures from abnormal brain areas and scores of clinical assessments were operated. Multivariate pattern analysis was applied to determine the most effective classification metrics among groups.
Cross-voxel correlation was lower in CSVD compared to healthy control. Abnormal brain regions were presented mainly in sensorimotor cortex, limbic/paralimbic system, and basal ganglia in CSVD. Notably, correlations between clinical assessment scores and NVC-related metrics in these areas were significant before correction. CBF/ALFF ratio exhibited superior classification performance between healthy control and CSVD with normal cognition, while a combination of dALFF and CBF effectively differentiated between CSVD patients with normal and impaired cognition.
Our investigation finds neurovascular decoupling using ATT-corrected CBF, dALFF and ALFF, as well as suggests effective classification metrics in CSVD with/without cognition impairment, potentially improving diagnostic and therapeutic strategies.
探讨脑小血管病（CSVD）中神经血管耦合（NVC）的改变、神经影像学指标与临床评估之间的关系以及分类指标。
根据CSVD量表和蒙特利尔认知评估，将参与者分为健康对照组、认知正常的CSVD组和认知障碍的CSVD组。结合经动脉通过时间调整的脑血流量（CBF）以及低频波动的动态/静态幅度（dALFF、ALFF）来评估NVC并确定组间差异。对异常脑区的测量指标与临床评估得分进行偏斯皮尔曼相关性分析。应用多变量模式分析来确定各组中最有效的分类指标。
与健康对照组相比，CSVD组的跨体素相关性较低。CSVD组的异常脑区主要出现在感觉运动皮层、边缘/边缘旁系统和基底神经节。值得注意的是，在这些区域，临床评估得分与NVC相关指标之间在校正前具有显著相关性。CBF/ALFF比值在健康对照组和认知正常的CSVD组之间表现出卓越的分类性能，而dALFF和CBF的组合能够有效区分认知正常和认知障碍的CSVD患者。
我们的研究发现使用经ATT校正的CBF、dALFF和ALFF存在神经血管解耦现象，同时提出了在有/无认知障碍的CSVD中有效的分类指标，这可能改善诊断和治疗策略。
