Yang Jiaojiao, Jiang Xiaofeng, Gu Lili, Li Jiahao, Wu Ying, Li Linghao, Xiong Jiaxin, Lv Huiting, Kuang Hongmei, Jiang Jian
Department of Radiology, The First Affiliated Hospital of Nanchang University, 17 Yongwaizheng Street, Nanchang 330006, China.
Neuroimaging Laboratory, Jiangxi Province Medical Imaging Research Institute, 17 Yongwaizheng Street, Nanchang 330006, China.
Brain Sci. 2023 Sep 22;13(10):1357. doi: 10.3390/brainsci13101357.
The purpose of this study was to explore the resting-state functional connectivity (FC) changes among the pain matrix and other brain regions in herpes zoster (HZ) and postherpetic neuralgia (PHN) patients. Fifty-four PHN patients, 52 HZ patients, and 54 healthy controls (HCs) underwent resting-state functional magnetic resonance imaging (rs-fMRI) scans. We used a seed-based FC approach to investigate whether HZ and PHN patients exhibited abnormal FC between the pain matrix and other brain regions compared to HCs. A random forest (RF) model was constructed to explore the feasibility of potential neuroimaging indicators to distinguish the two groups of patients. We found that PHN patients exhibited decreased FCs between the pain matrix and the putamen, superior temporal gyrus, middle frontal gyrus, middle cingulate gyrus, amygdala, precuneus, and supplementary motor area compared with HCs. Similar results were observed in HZ patients. The disease durations of PHN patients were negatively correlated with those aforementioned impaired FCs. The results of machine learning experiments showed that the RF model combined with FC features achieved a classification accuracy of 75%. Disrupted FC among the pain matrix and other regions in HZ and PHN patients may affect multiple dimensions of pain processing.
本研究旨在探讨带状疱疹(HZ)和带状疱疹后神经痛(PHN)患者疼痛矩阵与其他脑区之间静息态功能连接(FC)的变化。54例PHN患者、52例HZ患者和54例健康对照(HC)接受了静息态功能磁共振成像(rs-fMRI)扫描。我们采用基于种子的FC方法,研究与HC相比,HZ和PHN患者在疼痛矩阵与其他脑区之间是否表现出异常的FC。构建随机森林(RF)模型,探讨潜在神经影像指标区分两组患者的可行性。我们发现,与HC相比,PHN患者在疼痛矩阵与壳核、颞上回、额中回、扣带前回、杏仁核、楔前叶和辅助运动区之间的FC降低。HZ患者也观察到类似结果。PHN患者的病程与上述受损的FC呈负相关。机器学习实验结果表明,结合FC特征的RF模型分类准确率达到75%。HZ和PHN患者疼痛矩阵与其他区域之间FC的破坏可能会影响疼痛处理的多个维度。