影像科学与光化学2024，Vol.42Issue(4)：344-352,9.DOI:10.7517/issn.1674-0475.2024.04.07
乳腺结节恶性风险列线图预测模型的构建与验证
Construction and Validation of a Nomogram Prediction Model for Breast Nodule Malignancy Risk
摘要
Abstract
Objective:Based on multimodal ultrasound technology,the independent risk factors of breast cancer were analyzed,and a nomogram prediction model of malignant risk of breast nodules was constructed.The predictive value of the model was evaluated to guide clinical development of more effective screening strategies for breast cancer.Methods:Prospectively,we collected 228 patients with a total of 230 breast nodules from the Affiliated Hospital of North Sichuan Medical College from May 2021 to December 2023,including 146 in the benign group and 84 in the malignant group.The above nodules are classified into 3-5 categories by ultrasound BI-RADS and have been confirmed by pathology.The indicators were selected by univariate analysis,the receiver operating characteristic(ROC)curve was drawn and the area under the curve(AUC)was calculated to evaluate the diagnostic efficacy of each index.Better performing indicators were included in multivariate Logistic regression and constructed a normogram prediction model of malignant risk of breast nodules.The Bootstrap method was used to validate the model,and the consistency index(C-index),calibration curve,and decision curve were used to evaluate the predictive performance and clinical benefit rate of the model.Results:Univariate primary screening showed that the diagnostic efficacy of AP typing and RI for breast cancer was higher,and the AUC of both was statistically significant when compared with the AUC of CDFI grading,AP grading,and PSV(all P＜0.05).Among the SWE related parameters,Emax had the best diagnostic efficacy for breast cancer,which showed statistically significant differences compared with SWE typing and Emean's AUC(Z=2.742,3.174,all P＜0.05).Multivariate analysis showed that age,BI-RADS,RI,AP typing and Emax could be included as independent risk factors in the construction of breast cancer prediction nomogram model.The C-index of the nomogram was 0.996,and the corrected C-index for the internal validation of the model line was 0.992.The calibration curve showed a good agreement between the predicted and actual probabilities of the model.The decision curve showed that when the threshold probability of the model is between 0 and 1.0,patients can obtain positive net benefits.Conclusion:A multimodal ultrasound nomogram model based on age,BI-RADS,RI,AP typing,and Emax can effectively predict the benign and malignant nature of breast nodules,and has good clinical application value.The visualization of multimodal ultrasound nomogram can help guide clinical development of more reasonable breast cancer screening strategies,facilitate early detection and diagnosis of breast cancer,and improve the survival rate and quality of life of pa-tients.关键词
多模态超声/列线图/预测/乳腺结节/良恶性Key words
multimodal ultrasound/nomogram/prediction/breast nodules/benign and malignant分类
医药卫生引用本文复制引用
罗季平,官愫,岳文胜..乳腺结节恶性风险列线图预测模型的构建与验证[J].影像科学与光化学,2024,42(4):344-352,9.基金项目
川北医学院附属医院揭榜挂帅项目2022JB001 （）
