目的 应用Logistic回归模型分析颈动脉斑块新生血管的超声造影分级预测重度脑白质病变的价值.方法 对85例脑白质病变患者行超声造影检查并进行新生血管分级,应用Fazekas量表评分将入选者分为重度组和非重度组,应用Logistic回归模型分析重度脑白质病变的独立风险因素,并应用ROC曲线评估独立风险因素预测重度脑白质病变的价值.结果 重度组中高龄患者(＞70岁)比例明显高于非重度组(P=0.003),Ⅲ级和Ⅳ级颈动脉斑块患者发生重度脑白质病变的比例明显高于Ⅲ级和Ⅳ级颈动脉斑块患者(P=0.000),单因素和多因素Logistic回归分析显示年龄和颈动脉斑块新生血管超声造影分级均是重度脑白质病变的独立风险预测因素(P=0.044、0.000),OR值分别为3.404、4.899.ROC分析显示以新生血管分级＞Ⅱ级预测重度脑白质病变的灵敏度和特异度分别为70.27％、81.25％,其曲线下面积为0.755(95％CI:0.650～0.842)(P=0.000);以年龄＞70岁预测重度脑白质病变的灵敏度和特异度分别为75.68％、56.25％,其曲线下面积为0.660(95％CI:0.54～90.759)(P=0.0017);以颈动脉斑块新生血管分级＞Ⅱ级和年龄＞70岁预测重度脑白质病变的评估效能差异无统计学意义(Z=1.269,P=0.2045).结论 颈动脉斑块新生血管的超声造影分级是预测重度脑白质病变的独立风险因素.“,”Objective To evaluate the prediction of severe white matter lesions by the neovascularization grade of carotid plaque using contrast enhanced ultrasonography by Logistic regression model analysis.Methods 85 patients with white matter lesions were studied by contrast enhanced ultrasonography, then the neovascularization grade of carotid plaque was assessed in enrolled patients.The enrolled patients were divided into two groups according to Fazekas scores:severe group and non-severe group.The independent risk and predictive factors of severe white matter lesions were analysed by Logistic regression model and the utility of the prediction of severe white matter lesions by the independent risk and predictive factors was evaluated by ROC curve.Results The ratio of patients with advanced age (＞70 years old) in severe group is higher than that in non-severe group (P=0.003) and the ratio of severe white matter lesions in patients with Ⅲ and Ⅳ grade carotid artery plaque is higher than that in patients withⅠandⅡ grade carotid artery plaque (P=0.000).Age and the neovascularization grade of carotid plaque are both the risk and predictive factors of severe white matter lesions in single-factor and multiple-factor Logistic regression model, (P=0.044 and 0.000) and the ORvalue is 3.404 and 4.899, respectively.The sensitivity and specificity to prediction of severe white matter lesions by the neovascularization grade of carotid plaque (cutoff value＞Ⅱ) was 70.27％ and 81.25％, respectively and the area under ROC curve is 0.755 (95％ CI:0.650～0.842) (P=0.000).The sensitivity and specificity to prediction of severe white matter lesions by age (cutoff value＞70 years old) was 70.27％and 81.25％, respectively and the area under ROC curve is 0.755 (95％ CI:0.650～0.842) (P=0.000).There are no significant differences of prediction efficiency by between age and the neovascularization grade of carotid plaque (Z=1.269, P=0.2045).Conclusions The neovascularization grade of carotid plaques by contrast enhanced ultrasonography is the independent risk and predictive factor to prediction of severe white matter lesions.