Study On Prediction Of Axillary Lymph Node Metastasis Of Breast Cancer Based On Ultrasound Imaging Combined With Clinicopathological Features

Objective To evaluate he value of radiomics based on twodimensional ultrasound combined with clinicopathological factors in predicting axillary lymph node metastasis of breast cancer before operation.Methods A retrospective analysis of 160 patients with breast cancer who were treated in our hospital was performed.All of them were female patients.They were randomly divided into training group(110 cases)and verification group(50 cases)according to the ratio of 7:3.The patient’s age,tumor size,tumor location,tissue type,tissue grade,molecular classification,HER-2,ER,PR and Ki-67 clinicopathological factors were collected,and then the focus area of the ultrasound image was manually delineated by 3D Slicer 4.13.0 software to select(the region of interest,ROI)and extract features.Max-relevance and min-redundancy(MRMR)and Least absolute shrinkage and selection operator(LASSO)are used to reduce and screen the extracted features.Radiomics Score(RadScore)was constructed by LASSO algorithm,and the clinicopathological factors related to axillary lymph node metastasis were analyzed by using SPSS26.0 statistical software and R language,and then the joint model was established by combining Rad-Score with clinicopathological factors,and the nomogram was drawn.Using receiver operator characteristic curve(ROC)curve,the area under the receiver operating characteristic curve(AUC)was calculated,and the prediction efficiency of the model was evaluated by calibration curve and(Decision curve analysis,DCA).Results 837 features were extracted from the two-dimensional image of each breast lesion,and 11 related features were screened out to calculate Rad-Score.The tumor size and tissue grade were independent predictors in the clinicopathological features,and the AUC of the training group and the verification group in the imaging group model and the combined model were 0.769,0.699,0.841 and 0.831,respectively.In the calibration curve,both of them show good calibration degree,and the decision curve analysis shows that the joint model has better clinical practicability than the imaging model.Conclusion The radiomics score of two-dimensional ultrasound images,tumor size and tissue grade were independent predictors of axillary lymph node metastasis in breast cancer.In this study,the combined model based on radiomics score and clinicopathological features of twodimensional ultrasound images has better predictive performance than radiomics model.