Objective To explore the diagnostic value of double section ultrasound radiomics (maximum section of long axis and vertical section of short axis) in breast cancer. Methods Breast ultrasound data of 237 patients from two top-three hospitals, Affiliated Sir Run Run Shaw Hospital of Zhejiang university School of Medicine and Qujing No.1 Hospital, Affiliated Hospital of Kunming Medical University, were retrospectively collected from December 2022 to December 2023. Patients were randomly divided into a training set of 166 cases and a test set of 71 cases according to a ratio of 7:3. ROI (region of interest) was delineated on the maximum section of the long axis and the vertical section of the short axis, respectively. At the same time, automatic conformal expansion of 3 mm was performed to extract the image omics features of the double section. The clinical ultrasound feature model (model 1), the maximum long-axis section model (model 2), the short-axis vertical section model (model 3), the maximum long-axis section model + the short-axis vertical section model (model 4), the maximum long-axis section + the short-axis vertical section model + the clinical ultrasound feature model (model 5) were established respectively. Through the Receiver Operating Characteristics (ROC), Area Under Curve (AUC), Decision Curve Analysis (DCA) to evaluate the diagnostic efficiency of the model. Results For model 1, the AUC of training set was 0.89, and the AUC, sensitivity, specificity and accuracy of test set were 0.89, 0.88, 0.82and 0.85, respectively. For model 2, the training set AUC is 0.87, and the test set AUC, sensitivity, specificity and accuracy are 0.85, 0.78, 0.77 and 0.77, respectively. For model 3, the training set AUC is 0.90, and the test set AUC, sensitivity, specificity and accuracy are 0.90, 0.78, 0.90 and 0.85, respectively. For model 4, the training set AUC is 0.92, and the test set AUC, sensitivity, specificity and accuracy are 0.94, 0.72, 0.95 and 0.85, respectively. For model 5, the training set AUC is 0.98, and the test set AUC, sensitivity, specificity and accuracy are 0.94, 0.88, 0.95 and 0.92, respectively. Model 5 had the highest diagnostic efficiency, and the Delong test showed P < 0.05, the difference was statistically significant. Conclusion Based on the maximum long-axis section + short-axis vertical section + clinical ultrasound features (Model 5), the diagnosis efficiency is the best, which can better assist the sonographer to make decisions, effectively reduce the number of delayed treatment, reduce unnecessary overtreatment, and benefit patients.