Objective: To explore the diagnostic value of the diagnostic model based on the qualitative features of contrast-enhanced ultrasound and quantitative parameters of VueBox software in the diagnosis of breast BI-RADS 4 types of lesions. Methods: 142 patients with breast masses diagnosed by ultrasonography as BI-RADS type 4 lesions and confirmed by pathology were selected in our hospital. There were 54 cases in benign group and 88 cases in malignant group. Conventional ultrasound and contrast-enhanced ultrasound were performed. VueBox software was used to draw contrast time-intensity curves and 12 quantitative parameters of contrast-enhanced ultrasound were obtained. Including peak time (TTP), rise time (RT), clearance time (FT), peak intensity (PE), inflow phase ratio (WiR), outflow phase ratio (WoR), local mean transit time (mTTI), area under inflow phase curve (WiAUC), area under outflow phase curve (WoAUC), area under perfusion-clearance curve (W iWoAUC, inflow phase perfusion index (WiPI) and Area of interest (Area) were compared to compare the qualitative characteristics and quantitative parameters of CEUS between groups. Logisitic regression analysis was used to construct the diagnostic model of CEUS with qualitative characteristics and quantitative parameters. The working characteristic curves of subjects were drawn to compare and analyze the diagnostic efficiency of the two models in differentiating the four types of breast BI-RADS lesions. Results: Among the qualitative features of CEUS, there were statistical differences in enhancement intensity, enhancement time, enhancement range, enhancement direction, enhancement shape, contrast agent distribution, perfusion defect area, crab foot enhancement, lesion surrounding vessels and perforator vessels among groups (all P < 0.05). Among the quantitative parameters of contrast-enhanced ultrasound, there were statistically significant differences among RT, TTP and FT groups (all P < 0.05). The rise time, peak time and decline time of malignant lesions were shorter than those of benign lesions. Logistic regression diagnostic models P1 and P2 were constructed for qualitative characteristics and quantitative parameters, and the equations were Logit(P1)=-2.557+5.888X1-4.513X2+5.609X3, where X1 is reinforcement time,X2 is reinforcement intensity, and X3 is perforation-like vessel. Logit(P2)=-1.915+1.277X1-0.360X2-0.229X3, where X1 is RT,X2 is TTP, and X3 is FT. The area under ROC curve of the qualitative characteristics and quantitative parameters of CEUS for the diagnosis of the 4 types of benign and malignant lesions of BI-RADS was 0.931 (95%CI: 0.880-0.969) and 0.746 (95%CI: 0.663-0.817), respectively, with statistical difference (P=0.001). The diagnostic accuracy was 93.7% and 66.7%, the sensitivity was 0.988 and 0.470, and the specificity was 0.852 and 0.961, respectively. Conclusion: The qualitative and quantitative analysis of CEUS has certain value in the differential diagnosis of breast BI-RADS type 4 lesions. Ceus qualitative feature diagnosis model is superior to CEUS quantitative parameter diagnosis model, which can provide more optimized and accurate guidance for clinical diagnosis of breast BI-RADS type 4 lesions.