Diagnosis Of Axillary Lymph Node Metastasis In Breast Cancer Based On Radiomics

Breast cancer is a common malignant tumor in the female.Lymph node metastasis is easy to occur in the later stage of breast cancer,which affects the prognosis of breast cancer.Sentinel lymph node biopsy is currently the gold standard for the diagnosis of axillary lymph node metastasis,but this method has the disadvantages of radioactivity,time-consuming,complex operation,and a considerable proportion of complications.Therefore,imaging is still a commonly used detection method in clinical practice,but imaging has high requirements on the professional level of clinicians,and a long time of reading the images is prone to misdiagnosis and missed diagnosis.In view of the above problems,this paper established a rapid,accurate and non-invasive diagnostic method for axillary lymph node metastasis of breast cancer based on ultrasound images.The specific work contents and results are as follows:1.Diagnosis of benign and malignant breast nodules based on ultrasound images.First of all,the radiologist use the manual method to accurately classify the breast nodular lesion area,and used the feature extraction methods to extract the radiomics features of the lesion area.Secondly,different feature selection methods are used to select the radiomics features,compare the diagnostic efficacy of the optimal feature subset selected by different methods in each model,and evaluate the optimal feature selection method and the optimal classification model.Finally,the generalization performance of the optimal classification model of breast nodules was evaluated by means of cross validation.The results showed that the optimal feature subset based on the LASSO regression feature selection method performed best in the SVM classification model,and its accuracy,AUC,specificity and sensitivity were 0.9055,0.9072,0.8687 and0.9455,respectively.The model had good generalization performance.2.Ki-67 diagnosis of breast cancer based on ultrasound images of malignant nodules.First,the malignant breast nodules were included in the ultrasound images,the lesion areas were accurately divided,and the radiomics features were extracted.Secondly,different feature selection methods are used to select the radiomics features,compare the diagnostic efficacy of the optimal feature subset selected by different methods in each model,and evaluate the optimalfeature selection method and the optimal classification model.Finally,the generalization performance of the optimal model is evaluated by means of interactive validation.The results showed that the optimal feature subset based on the LASSO regression feature selection method performed best in the SVM classification model,with the accuracy,AUC,specificity and sensitivity of 0.8000,0.7893,0.8888 and 0.6896,respectively.The model had good generalization performance.3.Diagnosis of axillary lymph node metastasis of breast cancer based on ultrasound images of malignant nodules.First,the malignant breast nodules were included in the ultrasound images,the lesion areas were accurately divided,and the radiomics features were extracted.Secondly,the optimal feature subset was selected by different feature selection methods,and the logistic regression model was established based on different feature subsets.Meanwhile,ki-67 index was included to evaluate the diagnostic performance of the logistic regression model based on different feature subsets.Finally,the generalization performance of the best diagnostic model is evaluated by means of interactive validation.The results showed that the diagnostic efficiency of the logistic regression model based on the radiomics features and ki-67 index was significantly better than that based on the radiomics features.The accuracy,AUC,specificity and sensitivity of the model were 0.8160,0.8826,0.7543 and 0.8676,respectively.The model had good generalization performance.The results of this study indicate that the radiomics method based on ultrasound images is feasible for the diagnosis of axillary lymph node metastasis of breast cancer,and the inclusion of ki-67 index can improve the accuracy of the diagnosis of axillary lymph node metastasis.Compared with clinical diagnosis method,this method has the advantages of rapidity,noninvasiveness and accuracy.