DEVELOPING A MULTI-STAGE CAD TECHNIQUE FOR ROBUST AUTO DETECTION OF CALCIFICATION IN DIGITAL MAMMOGRAM

This paper presents multi-stage automatic technique to satisfy the needs early detection of breast classification. We focused on the preprocessing phase to improve the capability of detecting the calcification. We developed a universal filter, which is capable of removing all types of noise. Extraction of the breast region and delineation of the breast contour is an essential step in our technique. Primarily it allows the search for abnormalities to be limited to the region of the breast without undue influence from the background of the mammogram. Moreover, the results of the auto detection of the classification have been presented in 3D model to enable the phy sician to track the anomalism simply. We employ a precise segmentation technique to suppress artifacts and accentuate the breast region, followed by a detection algorithm to classify the breast tissue region. To demonstrate the capability of our multi-stage computer aided design (CAD) system it is extensively tested on mammograms from the Digital Database for Screening Mammography DDSM of the University of South Florida.