Current Development in Theory and Applications of Wavelets
Volume 1, Issue 1, Pages 1 - 24
(April 2007)
|
|
HIGH PERFORMANCE SYSTEM FOR DETECTING MICROCALCIFICATION CLUSTERS IN MAMMOGRAMS USING FUZZY LOGIC AND WAVELET
H. D. Cheng (U.S.A.), X. Y. Liu (U.S.A.), X. J. Shi (U.S.A.) and L. M. Hu (U.S.A.)
|
Abstract:
Breast cancer ranks second among cancer deaths in women. Primary prevention seems impossible since the causes of this disease still remain unknown. Early detection is essential in breast cancer control. Computer-aided mammography is an important and challenging task. An early sign of 30-50% of breast cancer cases detected mammographically is the appearance of clusters of fine, granular microcalcifications. In this paper, a novel high performance system to microcalcification detection, based on fuzzy logic and wavelet transformation is presented. First, we employ fuzzy entropy principle to fuzzify the images. Then, we use wavelet transformation to process the fuzzified image. A method using both local and global information is exploited to segment the microcalcifications, and a novel spatial influential function is applied to remove the isolated spots. Finally, the microcalcification clusters are detected and labeled. The free-response operating characteristic curve (FROC) is used to evaluate the performance. Comparing with existing algorithms using the same set of mammograms, the proposed approach achieves better results, producing both low FPs and low FNs rates, and it can detect microcalcifications even in the mammograms of very dense breasts. |
| Keywords and phrases: fuzzy logic, maximum entropy principle, wavelet transformation, microcalcifications, spatial influential function, contrast enhancement. |
| |
| Number of Downloads: 238 | Number of Views: 523 |
|
