Jeon-Hor Chen1,2, Muqing Lin1,
Fu-Ju Lei2, Jia-Pei Wu2, Siwa Chan3, Orhan
Nalcioglu1, Min-Ying L. Su1
1Center for Functional
Onco-Imaging & Department of Radiological Science, University of
California Irvine, Irvine, CA, United States; 2Department of Radiology, China Medical University
Hospital, Taichung, Taiwan; 3Department of Radiology, Taichung
Veterans General Hospital, Taichung, Taiwan
In general computer-aided algorithms for segmentation of breast density on MRI work well, but some errors may remain in small regions of the breast. Two major problems were the strong intensity inhomogeneity within a large area, and the low contrast between fibroglandular tissue and fatty tissue. In this study the operator noted the errors and classified them into three types. Several correction strategies were developed based on the nature of these errors and demonstrated satisfactory correction results. These processes to identify potential errors and correction strategies are important for developing a fully automated procedure for quantitative analysis of MRI-based density.