ABSTRACT Objective To analyze the contrast-enhanced ultrasound (CEUS) features of the benign and malignant breast lesions, to build a CEUS predictive model for exploring it’s value in the risk assessment of BI-RADS 4 breast lesions. Methods 174 patients with 180 lesions underwent contrast-enhanced ultrasound (CEUS) in the Affiliated hospital of nanjing university of Chinese medicine from September 2018 to March 2020 were selected. With final pathology results as the gold standard.The patients were divided into the model-Established group (91 cases with 94 lesions) and the model validation group (83 cases with 86 lesions).Univariate and multivariate Logistic regression analysis was carried out on the features of CEUS patterns in the model-Established group to construct a prediction model and draw ROC curve. Using final pathology results as the gold standard, the diagnostic efficacy of the CEUS predictive model was evaluated in the model validation group.Results Three independent variables of Logistic regression analysis were "crabfoot" or vascular distortion (OR=11.308, P < 0.001), enhanced homogeneity (OR=5.980, P=0.006) and increased lesion scope (OR=3.377, P=0.001). The CEUS predictive model was built as y=-4.239+2.425x8+1.788x3+1.217x5.The area under ROC curve of the CEUS predictive model in distinguishing between benign and malignant breast lesions was calculated to be 0.909, and the diagnostic sensitivity, specificity, accuracy, positive predictive value and negative predictive value of the CEUS predictive model were 95.83%, 81.58%, 89.53%, 86.79% and 93.94%, respectively.Conclusion The breast CEUS predictive model can predict malignant risk of breast lesions more accurately and has high risk evaluation of BI-RADS 4 breast lesions. KEY WORDS Ultrasonography;Contrast medium;Breast diseases;Logistic models.