Mo Jiajie, Yang Bowen, Wang Xiu, Zhang Jianguo, Hu Wenhan, Zhang Chao, Zhang Kai
Department of Neurosurgery, Beijing Tiantan Hospital, Capital Medical University, Beijing, China.
Department of Neurosurgery, Beijing Neurosurgical Institute, Capital Medical University, Beijing, China.
Ann Transl Med. 2022 Jul;10(13):741. doi: 10.21037/atm-22-630.
Surface-based cortical morphological patterns provide insight into the neural mechanisms of Parkinson's disease (PD). Explorations of the relationship between these patterns and the clinical assessment and treatment effects could be used to inform early intervention and treatment planning.
We recruited 78 PD patients who underwent presurgical evaluation and 55 healthy controls. We assessed neocortical sulcal depth, gyrification index, and fractal dimension and applied a general linear model using the multivariate Hotelling's -test to determine the joint effect of surface-based shape abnormalities in PD. The relationship between the neuroimaging pattern and clinical assessment was investigated using a multivariate linear regression model. A machine learning model based on surfaced-based features was used to predict responses to medication and deep brain stimulation (DBS).
The surface-based neuroimaging pattern of PD included decreases in morphological metrics in the gyrus (left: F=4.32; right: F=4.13), insular lobe (left: F=4.87; right: F=4.53), paracentral lobe (left: F=4.01; right: F=4.26), left posterior cingulate cortex (F=4.48), and left occipital lobe (F=4.27, P<0.01). This pattern was significantly associated with cognitive performance and motor symptoms (P<0.01). The machine learning model using morphological metrics was able to predict the drug response in the tremor score (R=-0.34, P<0.01) and postural instability and gait disorders score (R=0.24, P=0.04).
We identified the surface-based neuroimaging pattern associated with PD and explored its association with clinical assessment. Our findings suggest that these morphological indicators have potential value in informing personalized medicine and patient management.
基于表面的皮质形态学模式有助于深入了解帕金森病（PD）的神经机制。探索这些模式与临床评估及治疗效果之间的关系，可用于指导早期干预和治疗规划。
我们招募了78例接受术前评估的PD患者和55名健康对照者。我们评估了新皮质沟深度、脑回化指数和分形维数，并应用多元Hotelling's检验的一般线性模型来确定PD中基于表面的形状异常的联合效应。使用多元线性回归模型研究神经影像模式与临床评估之间的关系。基于表面特征的机器学习模型用于预测药物治疗和深部脑刺激（DBS）的反应。
PD的基于表面的神经影像模式包括脑回（左侧：F = 4.32；右侧：F = 4.13）、岛叶（左侧：F = 4.87；右侧：F = 4.53）、中央旁小叶（左侧：F = 4.01；右侧：F = 4.26）、左侧后扣带回皮质（F = 4.48）和左侧枕叶（F = 4.27，P<0.01）的形态学指标降低。这种模式与认知表现和运动症状显著相关（P<0.01）。使用形态学指标的机器学习模型能够预测震颤评分（R = -0.34，P<0.01）和姿势不稳及步态障碍评分（R = 0.24，P = 0.04）中的药物反应。
我们确定了与PD相关的基于表面的神经影像模式，并探索了其与临床评估的关联。我们的研究结果表明这些形态学指标在指导个性化医疗和患者管理方面具有潜在价值。
