常规超声联合超声造影对乳腺BI-RADS 4类小结节的诊断价值  

The value of conventional ultrasound combined with contrast-enhanced ultrasound in the diagnosis of BI-RADS category 4 small breast nodules

作　　者：钟兆明 唐丽娜[1] 王瑶琴 陈轶洁[1] 杜忠实[1] Zhong Zhaoming;Tang Lina;Wang Yaoqin;Chen Yijie;Du Zhongshi(Department of Ultrasound,Fujian Medical University Cancer Hospital,Fujian Cancer Hospital,Fuzhou 350014,China)

机构地区：[1]福建医科大学附属肿瘤医院福建省肿瘤医院超声科,福州350014

出　　处：《中华超声影像学杂志》2021年第11期955-960,共6页Chinese Journal of Ultrasonography

摘　　要：目的探讨常规超声(conventional ultrasound,US)联合超声造影(contrast-enhanced ultrasound,CEUS)对乳腺影像报告和数据系统(Breast Imaging Reporting and Data System,BI-RADS)4类小结节(最大径≤2 cm)的诊断价值。方法纳入2015年9月至2018年8月就诊于福建省肿瘤医院的175例患者的175个BI-RADS 4类且最大径≤2 cm的乳腺结节的US及CEUS检查资料,以上结节均取得粗针穿刺或手术病理结果,采用盲法对采集的超声图像及视频进行分析。采用Logistic逐步回归法分析恶性结节在超声影像中的优势比征象,根据OR值构建风险预测评分模型。并重新调整BI-RADS分类,采用ROC曲线比较调整前后的诊断效能。结果多因素Logistic回归分析结果显示,乳腺恶性小结节的优势比征象为边缘不清晰(OR=3.32,P=0.052)、肿块内钙化(OR=7.42,P=0.002)、结构扭曲(OR=38.58,P<0.001)、导管扩张(OR=0.01,P=0.010)、腋窝淋巴结可疑或异常(OR=10.92,P=0.003)、增强后病灶扩大(OR=3.38,P=0.040)、滋养血管(OR=10.79,P=0.006)、增强后边缘不清晰(OR=6.24,P=0.003)。当乳腺良恶性小结节的鉴别分值为3.5分时,ROC曲线下面积、敏感性、特异性、准确性为0.951、87.80%、89.20%、88.57%。重新调整BI-RADS分类后,以调整后的BI-RADS 4a类为活检阈值,活检率由100%降低至58.86%,癌症检出率由46.86%上升至75.73%,而漏诊的恶性风险为2.29%。BI-RADS分类调整前后的ROC曲线下面积分别为0.838和0.937。结论US联合CEUS可以提高乳腺BI-RADS 4类小结节良恶性鉴别的诊断效能,降低不必要的穿刺活检。Objective To evaluate the value of conventional ultrasound(US)combined with contrast-enhanced ultrasound(CEUS)in the diagnosis of Breast Imaging Reporting and Data System(BI-RADS)category 4 small(≤2 cm)breast nodules.Methods A total of 175 breast nodules in 175 patients from Fujian Cancer Hospital between September 2015 and August 2018 classified as BI-RADS category 4 breast nodules with maximum diameter≤2 cm were evaluated by US and CEUS examinations.All nodules were examined by core-needle biopsy or surgical pathology.The collected ultrasound images and videos were analyzed by blind method.Stepwise Logistic regression was used to analyze the odds ratio of malignant nodules in ultrasound images,and the risk prediction score model was constructed according to OR value.The BI-RADS category was readjusted,and the diagnostic efficiencies before and after adjustment were compared with the ROC curve.Results Multivariate Logistic regression analysis showed that the odds ratios of breast malignant nodules were non-circumscribed margin(OR=3.32,P=0.052),calcification in the mass(OR=7.42,P=0.002),architectural distortion(OR=38.58,P<0.001),ductal dilatation(OR=0.01,P=0.010),suspicious or abnormal axillary lymph nodes(OR=10.92,P=0.003),enlarged lesion scope(OR=3.38,P=0.040),penetrating vessels(OR=10.79,P=0.006),and non-circumscribed margin after enhancement(OR=6.24,P=0.003).When the cut-off value was 3.5,the area under ROC curve,sensitivity,specificity and accuracy were 0.951,87.80%,89.20%and 88.57%,respectively.After adjusting BI-RADS classification and taking the adjusted BI-RADS category 4a as the biopsy threshold,the biopsy rate decreased from 100%to 58.86%,the cancer detection rate increased from 46.86%to 75.73%,and the risk of missed diagnosis was 2.29%.The area under ROC curve before and after BI-RADS classification adjustment was 0.838 and 0.937,respectively.Conclusions US combined with CEUS can improve the diagnostic efficiency of BI-RADS category 4 small breast nodules and reduce unnecessary biopsy.

关 键 词：超声造影 常规超声 乳腺小结节 乳腺影像报告和数据系统 Logistic回归 

分 类 号：R445.1[医药卫生—影像医学与核医学] R737.9[医药卫生—诊断学]