Related Experiment Video
Updated: May 3, 2026
Tracking the Mammary Architectural Features and Detecting Breast Cancer with Magnetic Resonance Diffusion Tensor Imaging
Published on: December 15, 2014
Identification of breast contour for nipple segmentation in breast magnetic resonance images
Chih-Ying Gwo1, Allen Gwo2, Chia-Hung Wei3
1Department of Information Management, Chien Hsin University of Science and Technology, Taoyuan 320, Taiwan.
Purpose:
The purpose of this study is to develop a method to simulate the breast contour and segment the nipple in breast magnetic resonance images.
Methods:
This study first identifies the chest wall and removes the chest part from the breast MR images. Subsequently, the cleavage and its motion artifacts are removed, distinguishing the separate breasts, where the edge points are sampled for curve fitting. Next, a region growing method is applied to find the potential nipple region. Finally, the potential nipple region above the simulated curve can be removed in order to retain the original smooth contour.
Results:
The simulation methods can achieve the least root mean square error (RMSE) for certain cases. The proposed YBnd and (Dmin+Dmax)/2 methods are significant due toP = 0.000. The breast contour curve detected by the two proposed methods is closer than that determined by the edge detection method. The (Dmin+Dmax)/2 method can achieve the lowest RMSE of 1.1029 on average, while the edge detection method results in the highest RMSE of 6.5655. This is only slighter better than the comparison methods, which implies that the performance of these methods depends upon the conditions of the cases themselves. Under this method, the maximal Dice coefficient is 0.881, and the centroid difference is 0.36 pixels.
Conclusions:
The contributions of this study are twofold. First, a method was proposed to identify and segment the nipple in breast MR images. Second, a curve-fitting method was used to simulate the breast contour, allowing the breast to retain its original smooth shape.
