基于CTA参数的颅内小动脉瘤破裂风险预测模型建立与验证  

Establishment and Validation of a Risk Prediction Model for Intracranial Small Aneurysm Rupture Based on CTA Parameters

作　　者：蒲阳 母其文 郭志伟 唐雨露 PU Yang;MU Qi-wen;GUO Zhi-wei;TANG Yu-lu(Department of Imaging,Nanchong Central Hospital,the Second Clinical College of North Sichuan Medical College,Nanchong 637000,Sichuan Province,China)

机构地区：[1]南充市中心医院·川北医学院第二临床医学院影像科,四川南充637000

出　　处：《中国CT和MRI杂志》2024年第6期37-39,共3页Chinese Journal of CT and MRI

基　　金：川北医学院2022年度四川省基层卫生事业发展研究中心资助项目(SWFZ22-C-88)。

摘　　要：目的 建立基于CTA参数的颅内小动脉瘤破裂风险预测模型并进行内部验证。方法 选择2018年1月至2022年12月在我院行CTA检查的226例颅内动脉瘤患者进行回顾性分析。收集可能影响颅内小动脉瘤破裂相关因素及CTA检查指标。根据有无动脉瘤破裂将患者分为2组,比较2组一般资料与CTA扫描资料,以LASSO回归筛选变量,用Logistic回归建立模型,列线图进行可视化。结果 本研究纳入的226例患者中共有121例(53.53%)出现破裂。破裂组与未破裂组间在高血压病、脑血管病家族史、动脉瘤部位、动脉瘤血管位置、异常搏动点、瘤颈、AR、SR、流动角及子囊差异均具有统计学意义(P<0.05)。LASSO回归基础上行多因素Logistic回归分析结果显示:异常波动点、 AR、SR、流动角及子囊为颅内小动脉瘤破裂的独立影响因素(P<0.05)。ROC分析结果显示,该模型预测颅内小动脉瘤破裂的AUC为0.886[95%CI(0.844,0.928)]。H-L拟合优度检验结果显示,模型预测的颅内小动脉瘤破裂概率与实际概率比较,差异无统计学意义(P>0.05);预测曲线与标准曲线基本拟合。决策曲线分析结果显示:当该列线图模型预测颅内小动脉瘤破裂概率阈值为0.15-1.00时,患者的净受益率大于0。结论 颅内小动脉瘤破裂主要受异常波动点、AR、 SR等因素的影响,本研究列线图模型用于预测颅内小动脉瘤破裂风险,可用于指导临床决策的制订。Objective To establish a risk prediction model for intracranial small aneurysm rupture based on CTA parameters and conduct internal validation.Methods A retrospective analysis was conducted in 226patients with intracranial aneurysms who underwent CTA examination in our hospital from January 2018 to December 2022.Releva nt factors and CTA examination indicators that may affect the rupture of intracra nial small aneurysms were collected.Patients were divided into two groups(rupture group and non-rupture group) based on the presence or absence of aneu rysm rupture.The general data of the two groups were compared with the CTA scan data.Patients were divided into 2 groups based on the presence or absence of ruptured aneurysms,and general data and CTA scan data were compared between the 2 groups,with variables screened by LASSO regression,modelled by logistic regression,and visualised by nomograms.Results A total of 121(53.53%) of the 226 patients included in this study experienced rupture.There were statistically significant differences in hypertension,family history of cerebrovascular disease,aneurysm location,aneurysm vessel location,abnormal pulsation point,tumor neck,AR,SR,flow angle and ascus between the rupture group and the non-rupture group(P<0.05).The results of multivariate logistic regression analysis based on LASSO regression showed that abnormal fluctuation points,AR,SR,flow angle,and subcapsules were independent influencing factors for the rupture of intracranial small aneurysms(P<0.05).Establish a column chart model for predicting the rupture of intracra nial small aneurysms based on the results of multiple factor analysis.The ROC analysis results showed that the model predicted an AUC of 0.886 [95% CI(0.844,0.928)]for ruptured intracra nial small aneurysms.The results of the H-L goodness-of-fit test showed that the difference between the probability of rupture of small intracranial aneurysms predicted by the model and the actual probability was not statistically significant(P>0.05);the predicted curve

关 键 词：颅内小动脉瘤 CTA 多因素分析 列线图模型 

分 类 号：R732.21[医药卫生—肿瘤]