论文部分内容阅读
目的:探讨简易评分系统在乳腺小肿块超声BI-RADS分类中的应用价值。方法:回顾性分析217个经病理证实的最大直径≤1 cm的乳腺小肿块超声声像图特征,按形态、内部回声、后方回声、微钙化、纵横比、Alder血流分类6个描述指标评分,每项1分,计算各乳腺小肿块的总分,依据乳腺影像与数据报告系统第4版为标准,假设BI-RADS各类的分数界限进行统计分析。结果:217个乳腺小肿块中良性肿块185个,其中0分98个,1分73个,2分12个,3分1个,4分1个。恶性肿块32个,其中1分1个,2分9个,3分18个,4分3个,5分1个。ROC曲线下面积为0.974,最佳截点为1.5分。假设0分为BI-RADS 3类,1分为4a类,2分为4b类,3分为4c类,≥4分为5类,则BI-RADS各类的阳性预测值分别为0.0%、1.4%、42.8%、94.7%、80.0%。结论:简易评分系统可为乳腺小肿块超声诊断提供客观量化依据,对其BI-RADS分类有较好的帮助。
Objective: To investigate the value of simple scoring system in the classification of small breast ultrasound BI-RADS. Methods: The 217 ultrasonographic features of small breast masses with maximum diameter ≤1 cm confirmed by pathology were retrospectively analyzed. The morphological, internal echo, posterior echo, microcalcification, aspect ratio, Alder blood flow classification , For each 1 point, calculate the total score of each small breast lumps, according to the fourth edition of the Breast Imaging and Data Reporting System as a standard, assuming the BI-RADS various types of statistical limits for the boundaries. Results: There were 185 benign masses in 217 small breast masses, including 0 98, 1 73, 2 12, 3 1, 4 1. 32 malignant tumors, including 1 point 1, 2 points and 9 points, 3 points and 18 points, 4 points and 3 points, 5 points and 1 point. The area under the ROC curve was 0.974 and the best cutoff was 1.5 points. Assuming that 0 is divided into BI-RADS category 3, 1 is classified as category 4a, 2 is classified as category 4b, 3 is classified as category 4c, and ≥ 4 is classified as category 5, then the positive predictive value for each category of BI-RADS is 0.0% 1.4%, 42.8%, 94.7%, 80.0%. Conclusion: The simple scoring system can provide an objective and quantitative basis for the ultrasonic diagnosis of small breast masses, which is helpful to the classification of BI-RADS.