Purpose To establish and validate an ultrasound-based radiomics nomogram for predicting ipsilateral axillary lymph node metastasis in stage T1 breast cancer.Materials and Methods 443 stage T1 breast cancer patients in the First Affiliated Hospital of Wannan Medical College from January 2012 to June 2021 were retrospectively collected.All patients were randomly divided into training(n=310)and validation(n=133)group.ITK-SNAP was used to delineate the tumor margins,and Pyradiomics software was used to extract features.Image omics models and Rad-scores were constructed after feature screening.Clinical model,radiomics model and combined diagnostic models were developed,with the combined model's nomogram constructed.The models'predictive values were assessed via receiver operating characteristic curves.Results Multivariate Logistic regression analysis showed that the positive axillary ultrasonography,high echo halo and abundant internal blood supply were the independent risk factors of axillary lymph node metastasis.Then the clinical model was constructed,and imaging omics model was also constructed by feature screening.The combined model,which incorporated clinical and imaging features,demonstrated superior predictive performance.In the training group,the area under the curve for the combined model was 0.822,which was significantly higher than that of the clinical model(0.765)and radiomics model(0.723)(P=0.002 1,P=0.001 8).In the validation group,the area under the curve for the combined model was 0.846,outperforming the imaging omics model(0.686,P=0.001 8)and the clinical model(0.783),though the latter difference was not statistically significant(P=0.111 3).Conclusion Ultrasound-based radiomics combined diagnostic model effectively predicts ipsilateral axillary lymph node metastasis in stage T1 breast cancer,demonstrating high clinical predictive efficiency.
Breast neoplasmsUltrasonographyLymphatic metastasisRadiomicsNomogramsForecasting
NETL
NSTL
万方数据
