[en] Objective: To characterize the value of ADC and their changes (ΔADC%) before and during the neoadjuvant chemotherapy, and their value in predicting the pathologic response of breast cancer to neoadjuvant chemotherapy. Methods: Diffusion weighted imaging and dynamic contrast enhanced magnetic resonance examinations were performed in 35 women with breast cancer before treatment and after the second and the forth cycle of neoadjuvant chemotherapy. All patients were divided into MHR (major histological response) group and NMHR (non-major histological response) group according to the final pathologic response. One-way ANOVA and t test were used to compare the differences of cancer tissue and normal tissue. Rank sum test was used to compare the difference of ADC and Δ ADC% between two groups. ROC analysis of ADC values and Δ ADC% were used to identify which parameter could be used to predict the pathologic response to neoadjuvant chemotherapy. Results: After the forth cycle, the ADC value and Δ ADC% in MHR group [1.54 x 10^-3 mm²/s (1.01-1.88) x 10^-3 mm²/s and 42.70% (6.31%- 67.86%)] were significant higher than those in NMHR group [1.22 x 10^-3 mm²/s (0.83-1.83) x 10^-3 mm²/s, 14.88% (-28.87%-39.69%, P<0.05)]. However, the ADC value and Δ ADC% in the two groups were not different before treatment and after second cycle treatment: 1.15 x 10^-3 mm²/s(0.85-1.62) x 10^-3 mm²/s and 12.00% (-12.96%-30.65%)in MHR; 1.15 x 10^-3 mm²/s(0.94-1.60) x 10^-3 mm²/s and 10.74% (-20.80%-41.81%, P>0.05) in NMHR. Only the Δ ADC₃% showed high efficacy in predicting the pathologic response (area under ROC curve was 0.88). Conclusion: The rate of ADC change in late cycle could be used to predict the pathologic response of breast cancer to neoadjuvant chemotherapy. (authors)