Neoadjuvant chemotherapy (NAC) is the standard treatment performed for patients with locally advanced breast cancer ^[1]. However, its effectiveness is case dependent hence it is crucial to assess the lesion response at early stage of the treatment so that tailored treatment may be made. Dimensional measure of the tumor based on contrast enhanced MR imaging (CE-MRI) is usually only effective at a later stage of the treatment. DWI was shown to be a viable means for early stage treatment response^[2], however its measurement reflects mixed effects of molecular diffusion and microcirculation perfusion. On the other hand, intravoxel incoherent motion (IVIM) allows the effects diffusion and perfusion to be separated, and has been reported to be superior to conventional DWI in several diseases^[3]. In this work, the value of IVIM-derived parameters in pre-treatment prediction and early response monitoring to NAC in locally advanced breast cancer is investigated.A total of 36 patients with locally advanced breast cancer received multiple-b DWI with 12 b values ranging from 0 to 1000 sec/mm² at the baseline, out of which 28 patients repeated the scan after the second cycle of NAC. CE-MRI as well as other conventional MR scans were also performed. Consent forms were obtained prior to the scan. Subjects were divided into major histological response (MHR) and non-major histological response (NMHR) groups according to the surgical pathologic specimen. ROIs were manually defined on diffusion weighted image (b=800 s/mm²) by encompassing the lesion while avoiding necrotic, hemorrhagic and cystic areas. Comparisons were made using either Student’s t test or non-parametric test on: 1) IVIM Parameters (D, D*, f) as well as maximum diameter (MD) and volume(V) of the lesions prior to and after 2 cycles of NAC; 2) The changes of the above measurements at two time points in MHR and NMHR groups. The diagnostic performance of different parameters was estimated using the receiver operating characteristic curve analysis.Figure 1 illustrates the images of a typical patient prior to and two cycles of NAC, effects of the NAC can be easily assessed qualitatively. Quantitatively, measured parameters at the two time points for both MHR and NMHR are shown in Table 1 & Figure 2. Several observations can be made: prior to NAC, the f value of MHR group was significantly higher than that of NMHR (32.40% vs. 24.40%); after two cycles of NAC, the D value was significantly higher and the f value was significantly lower in MHR than those in NMHR; also, △D and △f were both higher in MHR than those in NMHR. Change of the parameters between MHR and NMHR are summarized in Table 2. The prediction performance of △D value was the highest (AUC=0.933). When the optimal cut-off was set at -0.261x10^-3mm²/s, the values for sensitivity, specificity, PPV and NPV were up to 90%, 88.9%, 81.8% and 94.1%, respectively (Table 3).In this study, IVIM derived parameters as well as dimensional parameters derived from CE MRI were measured and compared at different time points of the NAC treatment for patient groups with different level of responses. Several findings were obtained: the fraction of microcirculation f can be a potentially used for identifying MHR patients prior to the treatment; during early treatment, D and f were able differentiate the two groups whereas conventional dimensional measures failed; changes of D and f may also be used differentiate MHR from NMHR group; prediction performance of the change of D for distinguishing the MHR from NMHR was excellent after 2 cycles of NAC. Overall, it can be concluded IVIM-derived parameters, especially the D and f value, outperformed dimensional parameters from CE-MRI, and may be biomarker for the pre-treatment prediction and early response monitoring to NAC in locally advanced breast. Patients who had a higher baseline f value, a higher mid-treatment D value and a lower mid-treatment f value were observed to respond better to neoadjuvant chemotherapy.