〔Abstract〕 241 patients were included in this study. Each patient had a BI-RADS 4 breast lesion, and each breast lesion was examined by conventional ultrasound, color doppler ultrasound and ARFI. The resistance index, shear wave velocity(SWV, SWVlesionsand SWVglands) and their ratios(SWVlesions/SWVglands) were recorded. The diagnostic performance of each quantitative parameter was evaluated by the subject operation characteristic curve, and then the classification algorithm was used for classification analysis to construct the prediction model. Among the 241 breast masses, 140 cases were malignant and 101 cases were benign. In terms of quantitative indicators RI, SWVlesionsand SWVlesions/SWVglands, RI and SWVlesionswere included in the classification algorithm, and the depth of the prediction model included two branches(SWVlesions>or ≤3.795 m/s; if SWVlesions≤3.795 m/s, then RI≤0.620 or 0.6200.790, and if SWVlesions>3.795 m/s then RI≤0.710 or>0.710). The classification algorithm led to an AUC of 0.938, a sensitivity of 98.6%, and a specificity of 57.4%. The prediction model of ultrasonic quantitative parameters can significantly improve the accuracy in the diagnosis of benign and malignant BI-RADS 4 breast lesions, and can avoid the puncture biopsy of 24% breast lesions.