乳腺癌改良根治术后胸壁复发结节的超声图像特征及预测模型构建  

Sonographic characteristics of chest wall recurrent nodules after modified radical mastectomy and establishment of prediction model

作　　者：何惠珍 叶洁仪 何艳萍 黄伟俊[1] HE Huizhen;YE Jieyi;HE Yanping;HUANG Weijun(Department of Ultrasound Diagnosis and Treatment,the First People’s Hospital of Foshan,Guangdong 528000,China)

机构地区：[1]广东省佛山市第一人民医院超声诊疗中心,528000

出　　处：《临床超声医学杂志》2023年第2期101-104,共4页Journal of Clinical Ultrasound in Medicine

摘　　要：目的总结乳腺癌改良根治术后胸壁复发结节的超声图像特征,并构建预测模型,探讨其应用价值。方法选取我院102例乳腺癌改良根治术后胸壁结节患者,其中恶性结节57例(复发组),良性结节45例(良性组),比较两组超声图像特征;应用多因素Logistic回归分析筛选乳腺癌术后胸壁结节恶性风险预测因素,并建立预测模型,分析其预测胸壁结节恶性风险的诊断效能。结果两组结节最大径、数目、累及组织层次、边界、形态、钙化、周围组织回声变化、血流分级和间隔时间比较差异均有统计学意义(均P<0.05)。多因素Logistic回归分析显示,胸壁结节最大径、数目、边界、血流分级、间隔时间是乳腺癌术后胸壁结节恶性风险的独立影响因素(均P<0.05)。建立预测模型为:Logit(P)=-5.126+0.086X1+2.315X2+2.549X3+1.874X4+1.945X5(X1:病灶最大径,X2:病灶数目,X3:边界,X4:血流分级,X5:间隔时间),其预测胸壁结节恶性风险的灵敏度、特异度、准确率、曲线下面积及95%可信区间分别为91.2%、82.2%、87.3%、0.935(0.890~0.979)。结论基于超声图像特征构建的预测模型能有效评估乳腺癌改良根治术后胸壁结节恶性风险。Objective To summarize the sonographic characteristics of chest wall recurrent nodules after modifiedradical mastectomy,and explore the application value of prediction model.Methods One hundred and two breast cancerpatients with chest wall nodules after modified radical mastectomy in our hospital were selected,including 57 patients withmalignant nodules(recurrence group)and 45 patients with benign nodules(benign group).The sonographic characteristics werecompared between the two groups.Multi-factor Logistic regression analysis was used to analyze the predictive factors of themalignant risk of chest wall nodules after breast cancer surgery,and a predictive model was established to analyze its diagnosticefficacy in predicting the malignant risk of chest wall nodules.Results There were significant differences in maximum diameter,number,involved layers,border,shape,calcification,the echo change of surrounding tissue,blood flow grading andpostoperative intervals between the two group(all P<0.05).Multi-factor Logistic regression analysis showed that maximumdiameter,number,borders,blood flow grading and postoperative intervals were the independent influence factors for the risk ofmalignant nodules(all P<0.05).The prediction model was as followed:Logit(P)=-5.126+0.086X1+2.315X2+2.549X3+1.874X4+1.945X5(X1:maximum diameter,X2:number,X3:border,X4:blood flow grading,X5:postoperative intervals).The sensitivity,specificity,accuracy and area under curve for predicting malignant chest wall nodules were 91.2%,82.2%,87.3%and 0.935(95%CI:0.890~0.979),respectively.Conclusion Prediction model based on sonographic characteristics caneffectively evaluate the malignant risk of chest wall nodules after modified radical mastectomy.

关 键 词：超声检查 乳腺肿瘤 恶性 胸壁复发 预测模型 

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]