Dynamic contrast-enhancement (DCE) MRI and diffusion-weighted imaging (DWI) are widely used in differentiating benign from malignant breast lesions. However, previous studies have reported a low specificity on DCE-MRI^1,2, and meanwhile there is a higher specificity but lower sensitivity on DWI. The combined strategy of DCE-MRI and DWI provides more quantitative parameters of tumor properties. In addition, quantitative parameters based on histogram-metrical could reflect the information of tumor heterogeneity. The purpose of this study is to explore the diagnostic value of combined DCE- MRI and DWI based on histogram-metrical in discriminating breast masses.This study was approved by the local institutional review board. Patients:32 patients (13 benignancy and 19 malignancy) determined by needle biopsy or/and surgical pathology were included and performed by DWI and DCE-MRI. MR Protocols: All MR examinations were performed by a 3.0T MR scanner (GE, Signa HDxt, USA) with 16-channel dedicated breast coil. The protocol included EPI (TR/TE, 7800ms/82ms; slice thickness, 5mm; FOV, 320mm; matrix, 128×128;b-value 1000s/mm2) and VIBRANT sequence (TR/TE, 4.4ms/2.1ms; FA, 15o; slice thickness, 5mm; FOV, 350mm; matrix, 416×320). Data analysis:The mean and histogram metrics (skewness, kurtosis, median, variance, entropy and energy) of pharmacokinetic parameters (i.e. Ktrans and kep) and apparent diffusion coefficient (ADC) at transverse slice with maximal tumor diameter were calculated by using commercially Omni-Kinetics 2.0 and MATLAB software. Statistical analysis:The comparisons between benign and malignant group in the different parameters were used by Mann-Whitney U test. Receiver operating characteristic (ROC) curve analyses were conducted to assess the prediction accuracies of each metric in discriminating the benignancy and malignancy by using the commercial software (MedCalc 13.0, Mariakerke, Belgium). The combined diagnostic indexes were estimated by Logistic regression. The ROC curves of combined diagnostic indexes were used to assess the efficiency in identifying benign from malignant lesions. For all tests, p<0.05 was considered as statistically significant difference.Mean and histogram metrics (skewness) of ADC showed significantly differences between benign and malignant lesions (p_max=0.007). On DCE-MRI mean and histogram metrics (median and variance) of Ktrans, mean and 5 metrics (skewness, median, variance, entropy and energy) of kep showed significantly differences between benign and malignant masses (p_max=0.034). The sensitivity and specificity of DWI, Ktrans and kep were 94.7% and 92.3%, 78.9% and 61.5%, 89.5% and 92.3%, respectively. The sensitivity and specificity of combined DCE-MRI and DWI were 94.7% and 100%, respectively (Table 1); The area under the ROC curve of combined DCE-MRI and DWI, DWI, Ktrans and kep alone were 0.988, 0.976, 0.733 and 0.915 respectively (Figure 1).The ROC analysis showed that combined DCE-MRI and DWI have a good accuracy in identifying benign from malignant lesions, especially the specificity was 100%.According to our knowledge, three parameters (Ktrans, kep and ve) calculated from DCE-MRI can provide quantifying tumor blood flow, microvasculature and capillary permeability³. DWI yields information about possible changes in the biological environment. The quantitative parameter-ADC reflects the mean diffusion of water molecules⁴. Therefore, combined DCE-MRI and DWI show the potential pathophysiological process in the tumor issue. Our study obtain that the sensitivity and specificity of combined DCE-MRI and DWI are 94.7% and 100%. It was inconsistent with a recent study, which showed the sensitivity and specificity of combined DCE-MRI and DWI with 91.6% and 85.5%⁵. It maybe attribute to histogram analysis in this study which has the potential to improve specificity Thus, combined DCE-MRI and DWI based on the histogram-metrical features have greater prospect of clinical application.Combined DWI and DCE-MRI based on histogram-metrical analysis have superior efficacy than either DCE-MRI or DWI alone in discriminating breast masses. Moreover, the parameters based on histogram-metrical can offer additional features of tumor heterogeneity.