A Visual Analytics Approach to Diagnosis of Breast DCE-MRI Data

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) of the breast has become an important image modality for early breast cancer detection. In comparison to conventional X-ray mammography it exhibits a higher sensitivity, but only moderate specificity. To improve the specificity, and therefore the distinction of benign and malignant changes, the lesion’s heterogeneity and the lesion enhancement kinetics have to be evaluated. We present a visual analytics approach for breast tumors in DCE-MRI data that divides a tumor in different regions with different perfusion characteristics by employing a region merging method. The resulting region-based representation allows for evaluation of the tumor’s heterogeneity and the region-wise qualitative and quantitative evaluation of the enhancement kinetics. The analyses are combined with a glyph-based representation for a fast overview of the whole lesion. We tested our approach with seven well-chosen lesions and compared them to their histopathologic reports.