A Computer Aided Diagnosis (CAD) is useful to improve the diagnostic accuracy of malignancy microcalcifications in early stage of detecting breast cancer. The previous work that is support vector machine (SVM) based microcalcification detection outperformed one using a conventional classifier. But, that method has some problem that is to spend much computational time for microcalcification detection. In this paper, we propose a segmentation method that selects high probable region as microcalcification for reducing computational complexity. Experimental results showed that the proposed method is not only reducing processing time, but also improving detection result.
展开▼