湖北医药学院学报2025，Vol.44Issue(5)：561-567,7.DOI:10.13819/j.issn.2096-708X.2025.05.008
临床特征联合超声影像组学评估BI-RADS 4类乳腺结节性质的列线图模型开发和验证
Development and Verification of a Nomogram Model for Evaluating the Nature of BI-RADS 4 Nodules by Clinical Features Combined with Ultrasound Imaging
摘要
Abstract
Objective To develop a nomogram model to evaluate the nature of Breast Imaging Reporting and Data System(BI-RADS)4 nodules by clinical features and ultrasound imaging and to verify its efficiency in clinical application.Meth-ods A total of 100 patients with BI-RADS 4 nodules admitted to the Huangshan City Hospital of Traditional Chinese Medi-cine from March 2021 to March 2024 were divided into an establishment set and a verification set at a ratio of 7∶3.Based on the postoperative pathological examination,77 nodules of 70 cases in the establishment set were categorized into benign or malignant groups.The clinical data of both groups were compared,and the ultrasound imaging features were collected.Multivariate Logistic regression was employed to analyze the factors affecting the nature of breast nodules.A nomogram mod-el was constructed based on these factors,and its efficiency was assessed using the receiver operating characteristic(ROC)curve and clinical decision curve(DCA).Results A total of 77 nodules were detected in 70 patients in the establishment set,including 35 benign nodules and 42 malignant nodules.Fifteen ultrasound imaging features were extracted,comprising 1 histogram feature,2 texture features,and 13 wavelet features.Multivariate Logistic regression analysis indicated that the nature of breast nodules was the dependent variable,with independent risk factors including age≥60 years,family history of breast cancer,focus diameter of 20～30 mm,unclear boundary,blood flow grade 3,and Radscore≥1,and early menar-che(age＜13 years)was identified as a protective factor(P＜0.05).The ROC curve analysis showed an area under the curve(AUC)of 0.922 for the establishment set and 0.835 for the verification set.The Kolmogorov-Smirnov goodness-of-fit test indicated good calibration of the model(χ2=4.268,R2=0.917,P＞0.05).The DCA results demonstrated that the decision curves for both sets were superior to the All and None intervention strategies within a threshold range of 5%～92%.Conclusion The combination of clinical features and ultrasound imaging provides a reliable reference for evaluating BI-RADS 4 nodules.关键词
临床特征/超声影像组学/乳腺影像报告及数据系统/乳腺结节/列线图Key words
Clinical features/Ultrasound imaging/Breast Imaging Reporting and Data System(BI-RADS)/Breast nod-ules/Nomogram model引用本文复制引用
王瑾,廖淑霞,王宇洁,胡丽娜..临床特征联合超声影像组学评估BI-RADS 4类乳腺结节性质的列线图模型开发和验证[J].湖北医药学院学报,2025,44(5):561-567,7.基金项目
高等学校省级质量工程项目(2021cjrh046) （2021cjrh046）
