Objective: To evaluate the predictive value of a predictive model for the efficacy of neoadjuvant chemotherapy (NAC) in breast cancer by multimodal ultrasound assessment of tumor biological behavior combined with systemic immune inflammation index.Methods: 98 patients with breast cancer were selected as prospective cohort study subjects. According to the NAC Miller-Payne grading system, it was divided into an effective group (MP4-5 grade, n=62) and ineffective group (MP1-3 level, n=36). All patients underwent multimodal ultrasound examination (including two-dimensional color Doppler, elastography, and contrast-enhanced ultrasound) before NAC to evaluate tumor blood flow (Adler grade), stiffness strain ratio (SR), perfusion peak intensity (PI) and time to peak (TTP). Using a fully automated blood cell analyzer to detect and calculate the Systemic Inflammatory Index (SII) based on the absolute value parameters in the blood routine report, and performing Miller-Payne grading of surgical pathology after NAC. Establish a joint prediction model using multiple logistic regression (Forward LR method), ROC curve analysis, calculate AUC, sensitivity, and specificity, Delong test compared the differences in curves.Results: The Adler blood flow grading, strain rate ratio, and peak intensity of the effective group were significantly lower than those of the ineffective group (P<0.001). On the contrary, the peak time of the effective group was significantly higher than that of the ineffective group (P<0.001). The SII, neutrophil count, and platelet count of the effective group were significantly lower than those of the ineffective group (P<0.05). On the contrary, the lymphocyte count in the effective group was significantly higher than that in the ineffective group (P<0.001). Multivariate Logistic regression analysis showed that tumor diameter, Adler blood flow grading, SR, PI, TTP, and SII were independent predictors of treatment efficacy (P<0.05). After multivariable adjustment, SR, PI, and SII remained independent predictors of treatment efficacy (P<0.05). ROC curve analysis showed that SR, PI, and SII all had good predictive value for efficacy failure, and the combined model significantly improved predictive performance.Conclusion: Multimodal ultrasound is used to evaluate the biological characteristics of tumor and SII to reflect the immune and inflammatory state of the body. The combination of the two can better indicate the development trend of the disease and the chemosensitivity of breast cancer patients when judging their condition.