Logistic回归模型评估超声弹性成像对乳腺良恶性结节的诊断价值  

Evaluation the value of ultrasound elastography in the diagnosis of benign and malignant breast nodules:logistic regression analysis

作　　者：宋晓雨[1] 萧家芳[2] 柏刚[2] SONG Xiao-yu;XIAO Jia-fang;BAI Gang(Department of Ultrasound,Taihe Hospital,Hubei University of Medicine,Hubei 442000,China)

机构地区：[1]十堰市太和医院(湖北医药学院附属太和医院)超声医学科,湖北442000 [2]湖北医药学院,湖北442000

出　　处：《放射学实践》2021年第3期413-416,共4页Radiologic Practice

摘　　要：目的:采用logistic回归分析建立乳腺结节超声特征参数为变量的诊断模型,评价超声弹性成像对乳腺良恶性结节的鉴别诊断价值。方法:以术后病理结果为金标准,回顾性分析132例患者共185枚乳腺结节的超声声像图特点,甄别出有价值的超声参数,建立logistic回归模型,评价其诊断效能及预测能力。结果:二分类logistic回归评估发现患者年龄、乳腺肿块的弹性模量值、周围脂肪的弹性模量值、结节的方位、边界对研究对象诊断为乳腺恶性存在影响,所建立的logistic回归方程为:Logit(P)=-31.53+0.082×χ12+0.067×χ11+0.217×χ10+2.912×χ3+3.051×χ9,模型具有统计学意义(χ2=182.20,P<0.000),模型的诊断符合率为94.3%。结论:通过乳腺结节超声弹性成像特点建立的回归诊断模型在临床决策中具有指导价值。Objective:To establish a diagnostic model of breast nodules with characteristic ultrasound parameters as variables by logistic regression analysis and to evaluate the value of ultrasound elastography in the differential diagnosis of benign and malignant breast nodules.Methods:The ultrasonographic features of 185 breast nodules in 132 patients were retrospectively analyzed based on the pathological results,and valuable ultrasound parameters were identified.The logistic regression model was established to evaluate the diagnostic efficacy and predictive ability.Results:The binary logistic regression showed that age,the elastic modulus of breast masses,the elastic modulus of surrounding fat,location and boundary of nodules affected the diagnosis of breast cancer.The logistic regression equation was:Logit(P)=-31.53+0.082×χ12+0.067×χ11+0.217×χ10+2.912×χ3+3.051×χ9,The established logistic model had statistical significance(χ2=182.20,P<0.001).The diagnostic accuracy of the model was 94.3%.Conclusion:The regression diagnosis model based on ultrasound elastography characteristics of breast nodules has guiding value in clinical decision-making.

关 键 词：超声检查 乳腺结节 弹性成像 LOGISTIC回归 

分 类 号：R445.1[医药卫生—影像医学与核医学] R655.8[医药卫生—诊断学]