Study on Breast Mass Segmentation in Mammograms

Breast cancer is regarded as one of the most frequent mortality causes among women. It is very important to create a system to diagnose suspicious masses in mammograms for early breast cancer detection. In this paper, we propose an automatic breast mass segmentation method based on patch merging met...

Full description

Saved in:
Bibliographic Details
Published in2015 3rd International Conference on Computer, Information and Application pp. 22 - 25
Main Authors Shenghua Gu, Yao Ji, Yunjie Chen, Jin Wang, Jeong-Uk Kim
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.05.2015
Subjects
Breast cancer
Breast mass segmentation
Computers
Generalized mean
GHFCM
Hierarchical distance function
Image segmentation
Mammography
Merging
Patch merging
Online AccessGet full text

Cover

More Information
Summary:Breast cancer is regarded as one of the most frequent mortality causes among women. It is very important to create a system to diagnose suspicious masses in mammograms for early breast cancer detection. In this paper, we propose an automatic breast mass segmentation method based on patch merging method and generalized hierarchical Fuzzy C Means (GHFCM). The patch merging method is used to obtain the adaptive region of interest (ROI), while the GHFCM method which is able to overcome the drawbacks of effect of image noise and Euclidean distance FCM which is sensitive to outliers is used to obtain the precisely mass segmentation results. The new method is evaluated over Mini MIAS dataset. The segmentation performance from experimentations demonstrates that our method outperforms the other compared methods.
ISBN:9781467377713
1467377716
DOI:10.1109/CIA.2015.13