多模态CT相关参数对急性脑梗死患者侧支循环转态的评估价值分析  

Multimodal CT parameters in assessing collateral circulation status in acute cerebral infarction patients

作　　者：田臻 杨扬 李明超[2] TIAN Zhen;YANG Yang;LI Ming-chao(Department of Neurology,Huai′an Hospital Affiliated to Yangzhou University(the Fifth People′s Hospital of Huai′an),Huai′an 223300,Jiangsu,China;不详)

机构地区：[1]扬州大学附属淮安医院(淮安市第五人民医院)神经内科,江苏淮安223300 [2]淮安市第一人民医院神经内科,江苏淮安223300

出　　处：《广东医学》2025年第1期130-137,共8页Guangdong Medical Journal

基　　金：淮安市自然科学研究计划项目(淮科[2023]34号-42)。

摘　　要：目的分析多模态CT相关参数临床应用于急性脑梗死(acute cerebral infarction,ACI)患者侧支循环转态的评估价值。方法回顾性分析淮安市第五人民医院2020年1月至2022年8月采用多模态CT扫描成像技术进行检查的150例ACI患者的影像组学特征及临床资料,按7∶3比例分为建模集(n=105)和验证集(n=45)。建模集按侧支循环是否转归分为转归组(n=68)和失败组(n=37),比较两组CT灌注成像(computer technology perfusion,CTP)的相关参数以及血清转化生长因子-β(TGF-β)、同型半胱氨酸(Hcy)。Logistic回归分别建立影像组学模型、临床模型及影像组学模型与临床特征结合的联合模型,筛选出ACI患者侧支循环转态的独立影响因素,构建风险预测模型,绘制列线图并验证。结果单因素分析结果显示,患者脑血容量(CBV)、脑血流量(CBF)、平均通过时间(MTT)、达峰时间(TTP)、TGF-β、Hcy、合并高血压差异有统计学意义(P<0.05);与失败组比较,转归组CBV、CBF、MTT、TTP、TGF-β、合并高血压患者例数显著要高且Hcy要低(P<0.05);在建模集中,影像组学模型与临床特征结合的联合模型曲线下面积为0.978,95%CI为0.954~1.000,其敏感度为91.9%,特异度为95.6%;验证集中AUC为0.989,95%CI为0.966~1.000,其敏感度为100%,特异度为98.5%。联合模型对比影像组学模型和临床模型具有更高的临床预测价值。结论基于CBV、CBF、MTT、TTP、TGF-β、Hcy、合并高血压等指标构建出多模态CT影像组学模型与临床特征结合的联合模型具备更高的预测价值,可为ACI患者侧支循环转态提供较好的评估效能。Objective To analyze the clinical application value of multimodal CT parameters in assessing collateral circulation status in acute cerebral infarction(ACI)patients.Methods A retrospective analysis was conducted on the imaging features and clinical data of 150 ACI patients who underwent multimodal CT scanning at the Fifth People′s Hospital of Huaian from January 2020 to August 2022.The patients were divided into a modeling set(n=105)and a validation set(n=45)in a 7∶3 ratio.The modeling set was divided into two groups based on collateral circulation outcome:the successful collateral circulation group(n=68)and the failure group(n=37).The CT perfusion(CTP)parameters and serum biomarkers including transforming growth factor beta(TGF-β)and homocysteine(Hcy)were compared between the two groups.Logistic regression models were developed to establish imaging models,clinical models,and a combined model incorporating both imaging features and clinical characteristics.Independent factors influencing collateral circulation status were identified,and a prediction model was constructed and validated with a nomogram.Results Univariate analysis showed significant differences in cerebral blood volume(CBV),cerebral blood flow(CBF),mean transit time(MTT),time to peak(TTP),TGF-β,Hcy levels,and hypertension status between the two groups(P<0.05).The collateral circulation successful group had significantly higher CBV,CBF,MTT,TTP,TGF-βlevels,and hypertension incidence compared to the failure group,with lower Hcy levels(P<0.05).In the modeling set,the combined model incorporating imaging features and clinical characteristics had an area under the curve(AUC)of 0.978(95%CI:0.954-1.000),with a sensitivity of 91.9%and specificity of 95.6%.In the validation set,the AUC was 0.989(95%CI:0.966-1.000),with a sensitivity of 100%and specificity of 98.5%.The combined model had superior predictive value compared to the imaging model and clinical model alone.Conclusion The multimodal CT imaging model combined with clinical characteristics,incor

关 键 词：多模态CT CT灌注 脑血管成像 急性脑梗死 侧支循环 

分 类 号：R743[医药卫生—神经病学与精神病学] R445[医药卫生—临床医学]