Abstract:

Objective To explore the value of a radiomics model based on ultrasound imaging in predict⁃ ing the sentinel lymph node status of breast cancer before surgery. Methods A total of 194 patients with early ⁃ stage breast cancer were retrospectively analyzed，all patients underwent preoperative breast ultrasound examina⁃ tion. According to the order of examination time，the patients were divided into training group（n = 103）and vali⁃ dation group（n = 88）. Image J software was used to manually delineate the lesion area in the ultrasound image along the tumor boundary. Pyradiomics was used to extract 1 130 features from each lesion area，and the features were screened using three statistical methods. Finally，an ultrasound imaging radiomics model was built using a lo⁃ gistic regression model. To evaluate the performance and value of an ultrasound imaging radiomics model in predict⁃ ing sentinel lymph node status，the receive operating characteristic curve（ROC），calibration curve，and decision curve were used. Results To construct the ultrasound imaging radiomics model，6 key image features were cho⁃ sen. The area of under the ROC curve of the model in the training group and the validation group were 0.795（95% CI：0.708～0.882）and 0.784（95%CI：0.688～0.881），respectively. The calibration curve showed that the model had a good calibration in both the training and validation groups；Additionally，the decision curve analysis con⁃ firmed that the model could support clinical decision ⁃making. Conclusion Ultrasound ⁃ based imaging radiomics model is of great value in predicting the sentinel lymph node status of breast cancer before surgery.

Key words:

ultrasonic image, image omics, breast cancer, sentinel lymph node, prediction model