基于流体力学的表面应力预测颈动脉斑块破裂的回归分析

Regression analysis on prediction of carotid plaque rupture based on surface stress of fluid dynamics

目的探讨颈动脉斑块表面应力与斑块破裂的关系及预测价值。方法前瞻性收集2016年1月至2020年12月江苏大学附属宜兴医院收治的颈动脉粥样硬化斑块患者67例,所有患者在接受头颈部CT血管造影(CTA)后进行为期2年的随访,根据随访结果分为破裂组16例和未破裂组51例。以CTA图像构建血管三维模型并获得斑块形态学参数,应用计算流体力学量化斑块各部位的表面应力。比较2组患者各应力参数差异,用logistic回归分析颈动脉斑块破裂的独立影响因素,用ROC曲线分析斑块表面应力参数对斑块破裂的预测效能,用线性回归分析斑块表面应力参数的影响因素。结果与未破裂组比较,破裂组年龄、吸烟、低密度脂蛋白胆固醇(LDL-C)、斑块总体积、顶部表面应力和近端底部-顶部表面应力差(ΔPBTSS)显著升高(P<0.05,P<0.01)。多因素logistic回归分析显示,LDL-C(OR=1.855,95%CI:1.542~8.349,P=0.043)和ΔPBTSS(OR=1.041,95%CI:1.016~1.067,P=0.002)是斑块破裂的独立危险因素。ROC曲线分析显示,LDL-C和ΔPBTSS二者联合预测斑块破裂的曲线下面积为0.906(95%CI:0.819~0.992,P=0.000)。单因素线性回归分析显示,斑块垂直径(β=22.320,95%CI:10.607~34.033,P=0.000)、近端坡度(β=3.002,95%CI:1.188~4.815,P=0.002)与ΔPBTSS密切相关,且垂直径与近端坡度存在显著交互作用(β=4.117,95%CI:3.169~5.065,P=0.000)。结论斑块表面应力与斑块破裂密切相关,有助于预测颈动脉斑块破裂的发生。

Objective To explore the relationship between carotid plaque surface stress and plaque rupture and its predictive value.Methods Sixty-seven patients with carotid atherosclerotic plaques admitted to Affiliated Yixing Hospital of Jiangsu University from January 2016 to December 2020 were prospectively recruited in this study.All of them underwent head and neck CT angiography(CTA)during a 2-year follow-up period,and were categorized into ruptured group(16 patients)and unruptured group(51 patients)based on the follow-up results.The CTA images were used to construct 3D models of the vessels and obtain plaque morphological parameters,and computational fluid dynamics was applied to quantify the surface stresses at each site of the.plaque.'The differences of each stress parameter were compared between the 2 groups.Logistic regression analysis was employed to identify the independent influencing factors of carotid plaque rupture,ROC curve was plotted to analyze the predictive efficacy of plaque surface stress parameters on plaque rupture,and linear regression analysis was applied to determine the influencing fac-tors for plaque surface stress parameters.Results Older age,larger ratio of smoking history,elevated LDL-C level,and higher total plaque volume,top surface stress,and proximal bottom-top surface stress difference(OPBTSS)were observed in the ruptured group than the unruptured group(P<0.05,P<0.01).Multivariable logistic regression analysis showed that LDLC(OR=1.855,95%CI:1.542-8.349,P=0.043)and OPBTSS(OR=1.041,95%CI:1.016-1.067,P=0.002)were independent risk factors for plaque rupture.ROC curve analysis showed that the AUC value of combination of LDL-C and△PBTSS in predicting plaque rupture was 0.906(95%CI:0.819-0.992,P=0.000).Univariate linear regression analysis showed that plaque droop di-ameter(β=22.320,95%CI:10.607-34.033,P=0.000)and proximal slope(β=3.002,95%CI:1.188-4.815,P=0.002)were strongly associated with OPBTSS,and there was a significant in-teraction between them(β=4.117,95%CI:3.169-5.065,P=0.000).Conclusion Plaque surface stress is closely associated with plaque rupture,and is helpful in predicting the occurrence of carotid plaque rupture.