〔Abstract〕 This prospective observational study included 361 benign and malignant breast nodules from 258 women who had underwent breast grayscale ultrasound and had been confirmed pathologically. A total of 396 image features of lesion areas in ultrasonic images were extracted. The radiomic signature was developed using least absolute shrinkage and selection operator algorithms after feature selection using the minimum redundancy maximum relevance method. Receiver operating characteristic curve(ROC)and calibration curve were used to test the performance of the model. The area under the ROC curve(AUC), accuracy, sensitivity, specificity, Youden index for the training cohort were 0.84,0.761,0.840,0.715,0.603. And the AUC, accuracy, sensitivity, specificity, Youden index for the validation cohort were 0.84,0.716, 0.823, 0.654, 0.536. The diagnostic results of data from the radiomics model were basically consistent with the actual situation. Radiomics based on grayscale ultrasound performs well in the differentiation of malignant from benign breast nodules which effectively avoids the subjective diagnosis, and has good clinical application value.