ABSTRACT
PURPOSE: To test the split dynamic magnetic resonance imaging (MRI) technique in the assessment of breast masses in which high spatial resolution and dual-echo high temporal resolution data are acquired during a single bolus injection. MATERIALS AND METHODS: Forty-four women with breast masses were examined using split dynamic MRI. Quantitative analysis was performed with pharmacokinetic modeling on T1 -weighted images and estimation of maximum peak change in R2 * images (R2 *-peakenh ). High spatial resolution data were interpreted by two radiologists using the Breast Imaging Reporting and Data System (BI-RADS). Mann-Whitney tests were used to determine the parameters ability for establishing or excluding malignancy. For both readers, diagnostic accuracy, with and without information from the quantitative analysis, was determined using receiver operating characteristic (ROC) analysis, and evaluated using pairwise comparison of the areas under the ROC curve (Az ) and McNemar tests. RESULTS: Significant parameters for establishing or excluding malignancy were R2 *-peakenh (P < 0.001), plasma volume (P = 0.006), and time-to-peak enhancement (P = 0.003) showing an Az of 0.928 combined. For one out of the two readers, diagnostic accuracy was significantly improved when adding quantitative kinetic analysis to the BI-RADS score (P = 0.017). CONCLUSION: High temporal resolution T1 -weighted and R2 * dynamic information combined with BI-RADS interpretations improved the diagnostic performance in differentiating malignant from benign breast masses compared to BI-RADS interpretations alone.
