出血性脑卒中患者水肿体积进展建模研究

Study on Modeling of Edema Volume Progression in Patients with Hemorrhagic Stroke

摘要: 本文利用Gauss非线性拟合，将患者的水肿体积与每一次检查点的间隔时间联系起来，得到全部患者的水肿体积与时间进展的关系曲线。同时，通过FCM模糊聚类算法将所有患者依据个人特征划分为四个亚组，分别利用Gauss非线性拟合对每个亚组的散点图进行拟合，再根据拟合曲线分析每个亚组的特征。此外，利用主成分分析(PCA)探究不同疗法对患者水肿体积进展的影响，分析表明止血治疗对于水肿体积的影响显著，通过降颅压治疗也可以有效降低患者的水肿体积。最后探究水肿体积与血肿体积以及治疗方法之间的关系，首先通过PCA分析得到血肿体积与治疗方法之间的关联度，再联系水肿体积与治疗方法之间的关系，得到血肿体积与水肿体积是正相关，以及止血治疗、降颅压治疗对于降低水肿、血肿体积有着较好的治疗效果的结论。

Abstract: This article uses Gauss nonlinear fitting to connect the patient’s edema volume with the interval between each checkpoint, and obtains the relationship curve between the edema volume and time progression of all patients. At the same time, all patients were divided into four subgroups based on personal characteristics through the FCM fuzzy clustering algorithm. Gauss nonlinear fitting was used to fit the scatter plot of each subgroup, and then characteristics of each subgroup were analyzed based on the fitting curve. In addition, principal component analysis (PCA) was used to explore the impact of different treatments on the progression of edema volume in patients. The analysis showed that hemostatic treatment has a significant impact on edema volume, and intracranial pressure lowering treatment can also effectively reduce the patient’s edema volume. Finally, the relationship between edema volume, hematoma volume and treatment methods was explored. First, the correlation between hematoma volume and treatment methods was obtained through PCA analysis, and then the relationship between edema volume and treatment methods was combined to obtain the positive correlation between hematoma volume and edema volume, as well as the conclusion that hemostatic treatment and intracranial pressure lowering treatment have a good therapeutic effect on reducing edema and hematoma volume.