Abstract:

Objective To investigate the valuable characteristics of vulnerable carotid plaques in predicting ischemic stroke by multimode ultrasound, and evaluate the diagnostic value of multimode ultrasound. Methods A total of 92 patients who underwent carotid ultrasound examination at department of ultrasound of Beijing Tiantan hospital, capital medical university from August 2016 to October 2017 were included in this study. The multimode ultrasound included routine two-dimensional ultrasound, contrast-enhanced ultrasonography (CEUS) and superb microvascular imaging (SMI). A total of 105 plaques were diagnosed, and all patients were divided into symptomatic group and nonsymptomatic group according to clinical symptoms. The ultrasonographic characteristics of plaques and the patients’ clinical data were included in univariate analysis, and the statistically significant factors were then included in Logistic regression analysis. The receiver operating characteristic (ROC) curve was plotted to test the sensitivity and specificity of predictive diagnosis. Results Univariate analysis showed that smoking index, serum total cholesterol, low density lipoprotein, apolipoprotein B, signal enhancement on CEUS, neovascularization on SMI, plaque surface morphology on CEUS and SMI, calcified nodule location were statistically significant. The Logistic regression analysis showed that high enhancement on CEUS [odds ratio (OR) 9.610, 95% confidence interval (CI) 2.734-33.779, P <0.001], irregular plaque surface on SMI (OR 4.483, 95%CI 1.323-15.189, P =0.016) and calcified nodule on plaque surface (OR 7.206, 95%CI 1.609- 32.265, P =0.010) were independent risk factors of ischemic stroke. The area under the ROC curve, the sensitivity and specificity of predicting ischemic stroke for the three factors were 0.723, 75.8%, 67.4%; 0.657, 50.1%, 76.1%; 0.679, 40.7%, and 82.6%, respectively; and that of multifactorial diagnosis were 0.837, 90.5%, and 62.0%. Conclusion The characteristics of vulnerable plaques, such as high enhancement on CEUS, irregular plaque surface on SMI and calcified nodule on plaque surface are independent risk factors of ischemic stroke. Multimode ultrasound can better detect plaque characteristics, and thus increase the accuracy of predicting ischemic stroke.

Key words: Plaque vulnerability; Ultrasound; Multi-factor regression analysis

摘要：

目的 将二维超声、微血管成像技术、超声造影颈动脉斑块相关的声像图特征及临床卒中的危险因 素纳入Logistic回归，明确对缺血性卒中最有影响的斑块特征及危险因素，明确超声联合诊断效能。 

方法 选择2016年8月-2017年10月在首都医科大学附属北京天坛医院超声科行颈动脉超声检查显 示存在颈动脉斑块的患者92例，共105个斑块，依据临床诊断是否发生缺血性卒中症状分为症状组和 非症状组。记录斑块声像图特征及患者临床资料。将单因素分析有差异的危险因素纳入Logistic回归 分析，并绘制受试者工作特征（receiver operating characteristic，ROC）曲线。 

结果 单因素分析显示，吸烟指数、血清总胆固醇、低密度脂蛋白、载脂蛋白B、超声造影（contrastenhanced ultrasonography，CEUS）增强程度、CEUS显示斑块表面形态、超微血管成像（superb microvascular imaging，SMI）显示新生血管、SMI显示斑块表面形态、钙化结节位置在症状组和非症状 组的差异有统计学意义。Logistic回归显示，CEUS高增强[比值比（odds ratio，OR）9.610，95%可信区 间（confidence interval，CI）2.734～33.779，P＜0.001]、SMI 显示斑块表面不规则（OR 4.483，95%CI 1.323～15.189，P =0.016）、表面钙化结节（OR 7.206，95%CI 1.609～32.265，P =0.010）是预测缺血 性卒中的易损斑块特征。上述征象ROC曲线下面积、预测缺血性卒中的灵敏度、特异度分别为0.723、 75.8%、67.4%；0.657、50.1%、76.1%；0.679、40.7%、82.6%。多模态超声联合诊断预测缺血性卒中曲 线下面积、灵敏度、特异度分别为0.837、90.5%、62.0%。 

结论 CEUS高增强、SMI显示斑块表面不规则、表面钙化结节提示颈动脉粥样硬化斑块易损，是预 测缺血性卒中的独立危险因素。超声联合诊断对预测缺血性卒中准确性显著提高。

关键词: 斑块易损性; 超声; 多因素回归分析