Abstract
The accurate segmentation of breast in mammograms is critical for computer-aided diagnosis (CAD) of breast cancer. The challenges of breast segmentation come from tissue superimposition, low contrast and low visible breast boundary, noise areas and artifacts. In this research, a novel Automatic Breast Segmentation (ABS) approach was proposed. The experiment results showed that the proposed ABS approach can segment breast boundary reliably and accurately.
Library
Statistics