Breast Cancer Diagnosis: A Multiparametric Magnetic Resonance Imaging Model with Dynamic Contrast Enhanced and Diffusion Weighted Imaging

To develop a multiparametric MRI model incorporating the ACR BI-RADS recommended descriptors for DCE-MRI, T2-weighted and DW imaging biomarkers for accurate breast cancer diagnosis. A multivariate logistic regression analysis of multiparametric MRI data from 210 breast tumors was performed to determine parameters that jointly predicted malignancy. A multiparametric MRI model incorporating quantitative and qualitative for DCE-MRI [mass margins (p=0.0012), initial EH (p=0.422) and delayed enhancement (0.0065)] and DW imaging biomarkers [ADCmean (p=0.0031)] enables an accurate breast cancer diagnosis. Results indicate that to maximize diagnostic accuracy a multiparametric MRI approach with DWI and DCE sequences must be considered.