摘要
目的探究基于CT血管成像(CTA)的临床-影像组学联合模型对主动脉壁内血肿保守治疗短期预后的预测价值。资料与方法回顾性收集2017年1月—2022年12月南昌大学第二附属医院经CTA确诊的119例主动脉壁内血肿患者临床和影像资料,所有患者进行保守治疗,并均行多次胸腹部CTA随访。在CTA图像上手动逐层勾画主动脉(包括真腔和假腔)进行特征提取和影像组学分析。根据患者随访复查CTA情况,将患者分为预后不良47例和预后良好72例。按7∶3随机分为训练集84例和测试集35例。从临床特征中筛选壁内血肿短期进展的危险因素,构建临床模型。采用最小绝对收缩与选择算子算法选择最优影像组学特征。将筛选的临床预测因素与影像组学特征整合,构建临床-影像组学联合模型和列线图。采用受试者工作特征曲线评估模型性能。通过校准曲线评估准确性和可靠性。采用决策曲线分析模型的临床效用。结果预后不良与预后良好患者年龄和性别差异有统计学意义(t=2.608,χ^(2)=4.909,P<0.05),且年龄是独立危险因素(OR=0.962,P=0.040)。联合模型在训练集的曲线下面积为0.829,准确度、敏感度、特异度分别为73.8%、93.9%、60.8%,高于影像组学模型和临床模型的曲线下面积(0.812、0.677)。Delong检验显示,联合模型效能优于临床模型(Z=-2.610,P<0.05),影像组学模型与联合模型的效能差异无统计学意义(Z=-1.005,P>0.05)。校准曲线显示联合模型在训练集及测试集中的拟合优度良好。决策曲线显示影像组学模型和联合模型的临床效用均高于临床模型。结论基于CTA的临床-影像组学联合模型对主动脉壁内血肿保守治疗短期预后具有较好的预测效能。
Purpose To explore the predictive value of a combined clinical-radiomics model based on CT angiography(CTA)to assess the short-term prognosis of conservative treatment for aortic intramural hematoma.Materials and Methods The clinical and CTA imaging data of 119 patients with aortic intramural hematoma diagnosed by CTA in the Second Affiliated Hospital of Nanchang University from January 2017 to December 2022 were retrospectively collected.All patients underwent non-surgical treatment and repeated CTA follow-up.The aorta(including true lumen and false lumen)was manually delineated layer by layer on CTA images,and the feature extraction and radiomics analysis were performed.According to the results of follow-up CTA,the patients were divided into poor prognosis(47 cases)and good prognosis(72 cases).They were randomly divided into training set(n=84)and testing set(n=35)according to 7:3.Univariate and multivariate analysis were used to screen the risk factors for short-term progression of intramural hematoma from clinical characteristics,and a clinical model was constructed.Least absolute shrinkage and selection operator algorithm was used to select the best radiomics features.The selected clinical predictors were integrated with radiomics features to construct a clinical-radiomics combined model and nomogram.The receiver operating curve was used to evaluate the performance of the model.Accuracy and reliability were assessed by calibration curve evaluation.Decision curve analysis was used to evaluate the clinical utility of the model.Results There were significant differences in age and gender between patients with poor prognosis and those with good prognosis(t=2.608,χ^(2)=4.909,both P<0.05),and age was an independent risk factor(OR=0.962,P=0.040).The area under the curve value of the clinical-radiomics combined model in the training set was 0.829,the accuracy,sensitivity and specificity were 73.8%,93.9%and 60.8%,respectively,which were higher than the area under the curve values of the radiomics model and the clinical model(0.812 and 0.677,respectively).Delong test showed that the combined model was superior to the clinical model(Z=-2.610,P<0.05),but there was no significant difference in performance between the radiomics model and the combined model(Z=-1.005,P>0.05).The calibration curve indicated that the combined model had a good goodness of fit in the training set and the testing set.The decision curve showed that both the radiomics model and the combined model had higher clinical utility than the clinical model.Conclusion The combined clinical-radiomics model based on CTA has a good predictive power for the short-term prognosis of aortic intramural hematoma after conservative treatment.
作者
赵祯义
王靖芝
周淑丽
彭云
龚良庚
ZHAO Zhenyi;WANG Jingzhi;ZHOU Shuli;PENG Yun;GONG Lianggeng(Department of Radiology,the Second Affiliated Hospital,Jiangxi Medical College,Nanchang University,Nanchang 330006,China;Jiangxi Key Laboratory of Intelligent Medical Imaging,Nanchang 330006,China)
出处
《中国医学影像学杂志》
北大核心
2026年第3期324-329,共6页
Chinese Journal of Medical Imaging
基金
国家自然科学基金项目(82260342)。
关键词
主动脉壁内血肿
CT血管造影术
保守治疗
影像组学
列线图表
预后
预测
Aortic intramural hematoma
Computed tomography angiography
Conservative treatment
Radiomics
Nomograms
Prognosis
Forecasting
