脑CT预测脑出血脑卒中相关性肺炎预后的临床研究  

Study on the Prognosis of Cerebral Hemorrhage Stroke Associated Pneumonia Predicted by Brain CT

作　　者：尚亚丽 毛景鹰 SHANG Ya-li;MAO Jing-ying(Department of Emergency,Nanjing Tongren Hospital,Southeast University School Of Medicine,Nanjing 211106,Jiangsu Province,China)

机构地区：[1]东南大学医学院附属南京同仁医院急诊科,江苏南京211106

出　　处：《中国CT和MRI杂志》2025年第3期46-48,78,共4页Chinese Journal of CT and MRI

基　　金：2020年度江苏省中医药科技发展计划项目(YB2020044)。

摘　　要：目的分析脑CT预测脑出血脑卒中相关性肺炎(SAP)预后的价值。方法本研究开发了三种模型来预测SAP的发生。采用十重交叉验证对模型的性能进行评价。此外,通过统计分析构建了概率图,根据四种肺炎类型显示SAP患者中哪些脑区更容易受到血肿的影响。结果244名患者提取了35个特征,三种模型为Logistic回归、支持向量机和随机森林,模型AUC范围为0.77到0.82。概率图显示,在中、重度SAP患者中,ICH的分布在左右两个半球之间存在差异,与SAP关系更密切的脑结构包括左脉络丛、右脉络丛、右海马区和左海马区。此外,脑出血量平均值和最大值与SAP的严重程度成正比。结论根据脑CT扫描对肺炎的发展进行分类是有效的。Objective To explore the value of brain CT in predicting the prognosis of cerebral hemorrhage stroke associated pneumonia(SAP).Methods three models were developed in this study to predict the occurrence of SAP.Ten fold cross validation was used to evaluate the performance of the model.In addition,a probability map was constructed by statistical analysis to show which brain regions in SAP patients were more vulnerable to hematoma according to the four types of pneumonia.Results 35 features were extracted from 244 patients.The three models were logistic regression,support vector machine and random forest,and the AUC of the models ranged from 0.77 to 0.82.The probability map shows that in patients with moderate and severe sap,the distribution of ICH is different between the left and right hemispheres,and the brain structures more closely related to sap include the left choroid plexus,the right choroid plexus,the right hippocampus and the left hippocampus.In addition,the average and maximum cerebral hemorrhage volume are directly proportional to the severity of SAP.Conclusion it is effective to classify the development of pneumonia according to brain CT scan.

关 键 词：脑出血脑卒中 卒中相关性肺炎 脑CT 

分 类 号：R563.1[医药卫生—呼吸系统]