Meeting Banner
Abstract #3029

Symmetrical Analysis of Bilateral Breasts Based on 3D MRI

Peter T. Fwu1, Jeon-Hor. Chen1, 2, Siwa Chan3, Dah-Cherng Yeh4, Chin-Kai Chang2, Julian Kao1, Muqing Lin1, Orhan Nalcioglu1, Min-Ying L. Su1

1Center for Functional Onco-Imaging, Department of Radiological Sciences, University of California, Irvne, CA, United States; 2Department of Radiology, China Medical University Hospital, Taichung, Taiwan; 3Department of Radiology, Taichung Veterans General Hospital, Taichung, Taiwan; 4Department of Surgery, Taichung Veterans General Hospital, Taichung, Taiwan

We applied quantitative methods to investigate the symmetry of bilateral breasts. The volume, as well as the 3D morphology (Circularity, Convexity, Irregularity) and texture properties (gray-level co-occurrence matrix) of the segmented fibroglandular tissues on MRI were analyzed. A high correlation between the volume measured from the left and right breasts is observed. The correlation for morphology parameters are still high, and worse for texture parameters. The analysis of symmetry is particularly important for detection of non-mass-like lesions on breast MRI. The result may provide a reference dataset for developing intellectual computer-aided diagnostic systems for detecting abnormal lesions on MRI.

Abstract PDF E-Poster Video

Keywords

abnormalities abnormality adding address aided analyze analyzed analyzing asymmetric asymmetrical asymmetry available background besides bilateral boundary breast breasts calcifications central chin china circularity close coefficient commercially comparative compressed computer confounded consisted contrast convexity correlation dataset defined degree dense density dependent depicted despite detection determined developed development diagnosis difficulty distribution effective enhancements evaluate evaluation every fatty feature future general gray healthy help heterogeneous homogeneity hospital hull human identical importance important included incorporated index indicate individual inside intensity internal irregularity known lesion lesions like linear mammography matrix measured measures medical micro morphological morphologically morphology natural nature needed none object occurrence overall particularly pattern percent physics pixel position potential previous previously problem procedure projection quantitative radiological radiology reading reconstructed reflecting regression respectively scanner segmented selected shape slice software sphere subject subjects summarized surface surgery symmetrical symmetry system systems table texture thought tissue tissues true understanding unreliable utilizes variance variation vary veterans volume volumetric women