武警医学2025，Vol.36Issue(3)：200-204,5.
基于多参数磁共振指标的BI-RADS 4类乳腺结节良恶性预测模型构建
Construction of prediction model of type 4 benign and malignant breast nodules by BI-RADS based on multi-parameter magnetic resonance index
林琼真 1陈佳韬 1方欣欣 1罗爱芳1
作者信息
- 1. 363000 漳州,第909医院/厦门大学附属东南医院放射诊断科
- 折叠
摘要
Abstract
Objective To construct a prediction model of type 4 benign and malignant breast nodules by breast imaging repor-ting and data system(BI-RADS)based on multi-parameter magnetic resonance index.Methods A total of 135 lesions in 135 patients with type 4 breast nodules by BI-RADS treated in the 909th Hospital from January 2022 to January 2024 were retrospectively included,and they were divided into malignant group and benign group according to the nature of the lesions.Single-factor and multifactor meth-ods were used to evaluate the independent predictors of type 4 benign and malignant breast nodules by BI-RADS.A prediction models of type 4 benign and malignant breast nodules by BI-RADS was constructed and the prediction efficiency was analyzed.Results There were 87 benign lesions and 48 malignant lesions in all 135 lesions of 135 patients.There were statistically significant differences in age(t=-5.485),lesion enhancement type(χ2=33.170),and average kurtosis value(t=-5.772)between the two groups(P<0.05).logistic multivariate analysis confirmed that age(OR=1.169),average kurtosis value(OR=483.393)and lesion enhancement type(OR=15.437)could be used as independent predictors of benign or malignant breast nodules of type 4 BI-RADS(P<0.05).Age,le-sion enhancement type,mean kurtosis value and P-value prediction probability were used to predict the ROC curve of breast nodule,and the areas under the curve were 0.765,0.720,0.770 and 0.896,respectively.Conclusions Age,mean kurtosis value and lesion enhancement type are independent predictors of type 4 benign and malignant breast nodules by BI-RADS.The data model based on the above three factors has shown good efficiency in predicting the nature of breast nodules in patients.关键词
乳腺肿瘤/磁共振/乳腺影像报告和数据系统Key words
breast tumor/magnetic resonance/breast imaging reporting and data system分类
临床医学引用本文复制引用
林琼真,陈佳韬,方欣欣,罗爱芳..基于多参数磁共振指标的BI-RADS 4类乳腺结节良恶性预测模型构建[J].武警医学,2025,36(3):200-204,5.
