A similarity study of content-based image retrieval system for breast cancer using decision tree

Abstract-Breast cancer is one of the leading causes of death for women all over the world. In 2012, it was estimated that there would be approximately 226 870 newly diagnosed cases and 39 510 deaths in the United States.1 Treatment at an early stage is important in order to reduce mortality from breast cancer. Screening mammography has been accepted as the major modality to reduce breast cancer mortality due to the increased detection of early cancers.2 Despite the improvement in mammographic quality, however, the low sensitivity of screening mammography in dense breast remains a major limitation.3 Ultrasonography (US) has been shown to be an
useful modality for characterizing breast masses as malignant or benign.4, 5 Improved imaging techniques have enabled better characterization of sonographically visible breast lesions, allowing for less invasive management. In Ref. 5, a sensitivity of 98.4% and a specificity of 67.8% were achieved to distinguish 750 benign and malignant lesions using US. Taylor et al. showed that when sonography was combined with mammography in characterizing 761 breast masses, the specificity was improved from 51.4% to 63.8%, the positive predictive