乳腺超声造影预测模型的建立及其对乳腺良恶性病变诊断效能的分析

目的分析乳腺超声造影（CEUS）的增强特征,构建乳腺病变CEUS预测模型,探讨该模型对乳腺良恶性病变的诊断价值。 方法选取2016年6月至2018年8月于丽水市人民医院及浙江大学医学院附属第二医院就诊的乳腺病变患者192例共195个病灶,所有病灶均为常规超声检查BI-RADS分类4类及以上,均经穿刺活检或手术取得病理结果。将病例分为CEUS组120例共123个病灶,均经CEUS检查;CEUS＋动态增强磁共振（DCE-MRI）组72例共72个病灶,均接受CEUS及DCE-MRI检查。对CEUS组120例患者的CEUS模式特征进行单因素及多因素Logistic回归分析,筛选预测乳腺恶性病变的CEUS危险因素,并建立预测模型,绘制ROC曲线。以CEUS＋DCE-MRI组72例患者的病理结果为"金标准",分别计算CEUS预测模型与DCE-MRI对乳腺良恶性病变的诊断效能。 结果Logistic回归分析结果显示诊断乳腺恶性病灶的CEUS特征性表现为增强后病灶范围增大（OR=12.941,P=0.003）,"蟹足"征或血管扭曲缠绕（OR=7.553,P=0.009）,灌注缺损（OR=5.670,P=0.024）。建立的风险预测模型即Logistic回归方程为：Y=-4.108＋2.560X₆＋2.022X₇＋1.735X₈。该模型预测乳腺良恶性病灶的ROC曲线下面积为0.953。以穿刺或术后病理结果为"金标准",CEUS风险预测模型诊断乳腺良恶性病变的敏感度、特异度、阳性预测值、阴性预测值以及准确性分别为93.0%、73.3%、93.0%、73.3%、88.9%;DCE-MRI诊断乳腺良恶性病变的敏感度、特异度、阳性预测值、阴性预测值以及准确性分别为94.7%、73.3%、93.1%、78.6%、90.3%。CEUS风险预测模型与DCE-MRI诊断乳腺良恶性病变的一致性较高（Kappa值=0.70）。 结论乳腺CEUS预测模型对鉴别良恶性病灶具有较高的诊断效能,且操作相对简单、检查时间短、可重复性好、价格相对低廉,不失为诊断乳腺良恶性病变的一种有效和可靠方法。

Comparative study of a contrast-enhanced ultrasound predictive model and dynamic contrast-enhanced magnetic resonance imaging in diagnosis of breast lesions

ObjectiveTo identify the contrast-enhanced ultrasound (CEUS) features of the breast, to build a CEUS predictive model for breast lesions, and to evaluate the diagnostic value of this model in distinguishing breast benign from malignant lesions. MethodsA total of 192 patients with 195 breast lesions were included in the study. The lesions were divided into two groups: 123 lesions in a CEUS group and 72 in a CEUS+ dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) group. In the CEUS group, CEUS were used to examine each lesion. Then, risky CEUS patterns in breast malignant lesions were identified by logistic regression analysis to build a breast CEUS predictive model. Using final pathology results as the gold standard, the diagnostic efficiencies of the CEUS prediction model and DCE-MRI were evaluated. ResultsThree independent variables, namely, increased lesion scope (OR=12.941), ″crab foot″ sign (OR=7.553), and filling defect (OR=5.670), were selected in the final step of the logistic regression analysis in the CEUS group. The CEUS predictive model was built as Y=-4.108+ 2.560X₆+ 2.022X₇+ 1.735X₈. Using final pathology results as the gold standard, the area under ROC curve of the CEUS predictive model in distinguishing between benign and malignant breast lesions was calculated to be 0.953, and the diagnostic sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of the CEUS predictive model were 93.0%, 73.3%, 93.0%, 73.3%, and 88.9%, respectively; the corresponding values of DCE-MRI were 94.7%, 73.3%, 93.1%, 78.6%, and 90.3%. The consistency between the CEUS risk prediction model and DCE-MRI in the diagnosis of benign and malignant breast lesions was high (Kappa value=0.70). ConclusionThe breast CEUS predictive model built here can predict the malignant risk of breast lesions more accurately. It is an effective and reliable method for the diagnosis of benign and malignant breast lesions because of its simple operation, short examination time, reproducibility, and relatively low price.