基于CCTA定量参数及影像组学在评估冠状动脉粥样硬化进展及预后中的应用研究

【作者】 陈瑞；

【导师】 祝因苏；

【作者基本信息】 南京医科大学 ， 放射影像学（影像诊断）（专业学位）， 2024， 硕士

【摘要】 第一部分基于CCTA冠周脂肪影像组学在冠脉粥样硬化斑块进展中的应用研究目的探讨基于冠状动脉计算机断层扫描血管成像(CCTA)的冠周脂肪组织(PCAT)影像组学特征在评估冠脉粥样硬化斑块进展中的潜在价值。材料和方法纳入2009年1月至2020年12月间在南京医科大学第一附属医院接受至少两次CCTA检查且两次CCTA检查间隔不小于2年的500名疑似或已知冠状动脉疾病(CAD)患者的临床及CCTA资料。患者按照7:3随机分成训练集(n=350)及测试集(n=150),根据斑块进展情况将患者分为进展组及非进展组,斑块进展定义为≥总斑块负荷(TPB)年变化率(ΔTPB/y)的中位数。利用随机森林方法优化PCAT影像组学模型,获得与斑块进展相关特征的重要性Radscore。使用单因素及多因素Logistic回归分析探讨斑块定量参数及PCAT衰减对斑块进展的影响。通过受试者工作特征曲线(ROC)评估斑块定量参数(Model 1)、冠周脂肪影像组学(Model 2)及二者联合(Model 3)模型预测斑块进展的效能。结果多因素回归分析显示,斑块定量参数非钙化斑块体积(NCPV)(OR=1.005,95%CI:1.001～1.009,P=0.027)和PCAT衰减(OR=1.037,95%CI:1.010～1.064,P=0.007)是冠脉粥样硬化斑块进展的独立预测因子。PCAT影像组学在训练集和测试集中预测斑块进展的效能均显著优于斑块定量参数(训练集AUC:0.814vs.0.615,P<0.001;测试集AUC:0.736 vs.0.594,P=0.007)。结论基于CCTA的斑块定量参数NCPV和PCAT衰减是冠脉粥样硬化斑块进展的独立预测因子,PCAT影像组学预测冠脉斑块进展的效能明显优于斑块定量参数。第二部分基于CCTA冠状动脉粥样硬化斑块定量参数在预测非阻塞性冠心病预后中的应用研究目的探讨基于冠状动脉计算机断层扫描血管成像(CCTA)斑块定量参数在预测非阻塞性冠心病患者发生主要心血管不良事件(MACE)中的价值。材料和方法纳入2010年9月至2022年9月在南京医科大学第一附属医院行CCTA检查的非阻塞性冠心病患者的临床、影像资料及预后(MACE)信息。首先,根据是否发生MACE将患者分为MACE(+)和MACE(-)组,并比较临床资料、斑块基线及进展定量参数在两组之间的差异。然后,采用单因素及多因素Cox回归分析,筛选出能有效预测患者发生MACE的影响因素。最后,利用斑块基线参数、斑块进展参数及二者联合分别构建预测模型,采用一致性指数-时间曲线评估模型预测效能。结果共有271例患者纳入研究,其中75例(27.6%)患者在随访期间发生MACE。与MACE(-)组相比,斑块基线参数模型DS(HR=1.030,95%CI:1.010~1.050,P=0.003)、TPB(HR=1.043,95%CI:1.013~1.073,P=0.004)、HRP(HR=2.619,95%CI:1.380~4.969,P=0.003)、FAI(HR=1.051,95%CI:1.008~1.096,P=0.020)是MACE的预测因素,斑块进展参数模型中ΔDS/y(HR=1.073,95%CI:1.031~1.117,P<0.001)、非梗阻进展为梗阻(HR=10.652,95%CI:6.055~18.740,P<0.001)可预测MACE的发生,二者联合模型中DS(HR=1.020,95%CI:1.000~1.040,P=0.047)、ΔDS/y(HR=1.086,95%CI:1.043~1.130,P<0.001)、非梗阻进展为梗阻(HR=8.002,95%CI:4.249~15.070,P<0.001)是MACE发生的独立预测因素。一致性指数-时间曲线结果显示,二者联合模型对非阻塞性冠心病患者发生MACE的预测效能优于斑块基线参数模型和斑块进展参数模型。结论基于CCTA的斑块定量尤其是进展参数在非阻塞性冠心病患者中可以预测发生MACE高风险人群,具有较好风险分层价值。更多还原

【Abstract】 Part 1 Radiomics Analysis of Peri-coronary Adipose Tissue from Baseline CCTA Enables Prediction of Coronary Plaque ProgressionObjective The relationship between plaque progression and peri-coronary adipose tissue(PCAT)radiomics has not been comprehensively evaluated.We aim to predict plaque progression with PCAT radiomics features and evaluate their incremental value over quantitative plaque characteristics.Materials and Methods Between January 2009 and December 2020,500 patients with suspected or known coronary artery disease(CAD)who underwent serial CCTA≥2 years apart were retrospectively analyzed and randomly stratified into a training and testing dataset with a ratio of 7:3.Plaque progression was defined with annual change in plaque burden exceeding the median value in the entire cohort.Quantitative plaque characteristics and PCAT radiomics features were extracted from baseline CCTA.Then we built three models including quantitative plaque characteristics(Model 1),PCAT radiomics features(Model 2)and the combined model(Model 3)to compare the prediction performance evaluated by AUC.Results The quantitative plaque characteristics of training set showed the values of noncalcified plaque volume(NCPV),fibrous plaque volume(FPV),lesion length(LL),PCAT attenuation in the plaque progression group exhibited significantly higher values compared to the non-progression group(P<0.05 for all).In multivariable logistic analysis,NCPV and PCAT attenuation emerged as predictors of coronary plaque progression independently.PCAT radiomics exhibited significantly superior prediction over quantitative plaque characteristics both in the training(AUC0.814 vs.0.615,P<0.001)and testing(0.736 vs.0.594,P=0.007)datasets.Conclusions NCPV and PCAT attenuation were identified as standalone predictors of coronary plaque progression.Furthermore,PCAT radiomics derived from baseline CCTA yielded significantly superior predictive capabilities compared to quantitative plaque characteristics alone.Part 2 Coronary Artery Plaque Progression in Predicting Prognosis of Non-Obstructive Coronary Artery DiseaseObjective To explore the value of coronary artery plaque progression based on coronary CT angiography(CCTA)in predicting the occurrence of major adverse cardiovascular events(MACE)in patients with non-obstructive coronary artery disease.Materials and methods The study included clinical,imaging,and prognosis(MACE)parameters of non-obstructive coronary artery disease patients who underwent CCTA at the First Affiliated Hospital of Nanjing Medical University from September 2010 to September 2022.Initially,patients were grouped based on the occurrence of MACE,and differences in clinical data,plaque baseline and progression parameters between the two groups were compared.Subsequently,univariate and multivariate Cox regression analysis were employed to identify factors that could effectively predict the occurrence of MACE in patients.Finally,models were constructed using plaque baseline parameters,plaque progression parameters,and combination of both.The concordance index-time curve was used to evaluate the risk stratification ability of the models.Results A total of 271 patients were included,of whom 75 cases(27.6%)experienced MACE during the follow-up period.In comparison to the MACE(-)group,plaque baseline parameters models including DS(HR=1.030,95%CI: 1.0101.050,P=0.003),TPB(HR=1.043,95%CI: 1.013~1.073,P=0.004),HRP(HR=2.619,95%CI:1.380~4.969,P=0.003),and FAI(HR=1.051,95%CI: 1.008~1.096,P=0.020)were predictive factors for MACE.In plaque progression models,ΔDS/y(HR=1.073,95%CI: 1.031~1.117,P<0.001)and progression from non-obstructive to obstructive(HR=10.652,95%CI: 6.055~18.740,P<0.001)were predictive of MACE occurrence.In the combined model,DS(HR=1.020,95%CI: 1.000~1.040,P=0.047),ΔDS/y(HR=1.086,95%CI: 1.043~1.130,P<0.001),and progression from non-obstructive to obstructive(HR=8.002,95%CI: 4.249~15.070,P<0.001)were independent predictive factors for MACE occurrence.Concordance index-time curve results indicated that the combined model had a better predictive efficacy for MACE in patients with non-obstructive coronary artery disease compared to models based on plaque baseline parameters and plaque progression parameters.Conclusions The plaque progression parameters based on CCTA have the potential to predict the high-risk population for MACE in patients with non-obstructive coronary artery disease,demonstrating good risk stratification value.更多还原