超声影像学报告及数据系统鉴别乳腺肿块良恶性的多因素分析  

The Multi-factor Analysis of Breast Imaging Reports and Data Systems of Ultrasound for Differentiating Benign from Malignant Masses

作　　者：周玮珺[1] 孔文韬[1] 张捷[1] 吴敏[1] 

机构地区：[1]南京鼓楼医院,江苏南京210008

出　　处：《肿瘤学杂志》2016年第4期265-268,共4页Journal of Chinese Oncology

摘　　要：[目的]分析乳腺超声影像学报告及数据系统中各个特征诊断乳腺良恶性肿瘤的价值。[方法]分析103个肿块的形态、方位、边缘、内部及后方回声、钙化、内部血供、腋下淋巴结情况,以病理结果为金标准,对各特征进行单因素和多因素分析,通过ROC曲线评价诊断结果。[结果]单因素分析显示形态不规则、边缘不光整、内部回声不均匀、后方回声衰减、微钙化、腋下淋巴结形态异常提示肿块恶性程度高(P<0.05)。多因素分析显示形态和边缘是诊断良恶性最相关的独立因素。形态不规则和边缘不光整诊断乳腺恶性肿瘤灵敏度及特异性分别为85.1%、78.6%和85.1%、80.4%。[结论]肿块形态及边缘是诊断肿块良恶性最敏感的特征。结合内部回声、后方回声、钙化、内部血供及腋下淋巴结能更准确地鉴别乳腺肿块良恶性。[Objective] To analyze the value of ultrasonic features of breast imaging reporting and data system（BI-RADS） on diagnosing the benign and malignant breast tumors. [Methods] Ultrasonic features including shape,orientation,margin,inner echo,posterior echo,calcification,internal blood supply and axillary lymph nodes of 103 breast masses were analyzed. With pathologic results as the golden standard,the single factor analysis and multifactor analysis of the various characteristics of breast masses were carried. Receiver-operating characteristic curve（ROC curve）was drawn. [Results] The single factor analysis showed irregular shape,not circumscribed margin,heterogeneous echo,posterior acoustic shadow,microcalcification and abnormal axillary lymph node were highly correlated with malignant masses（P0.05）. The multiple factors analysis showed the shape and margin were the independent factors for differentiating benign from malignant breast masses. The sensitivity and specificity of the shape and margin for diagnosis of breast ma-lignant masses were 85.1%,78.6% and 85.1%,80.4% respectively. [Conclusions] The shape and margin are the most sensitive characteristic for differentiating benign from malignant breast masses.Combined with the internal echo,posterior echo,calcification and internal vascularity and axillary lymph nodes,we can more accurately identify breast benign and malignant masses.

关 键 词：乳腺肿瘤 影像学报告及数据系统 超声 

分 类 号：R737.9[医药卫生—肿瘤]