A Novel Image Enhancement Scheme for Computer Aided Diagnosis of Breast Cancer using Fully Digital Mammography

The purpose of this paper is to develop a imageenhancement system for computer-aided diagnosis (CAD) of breast cancer using full digital mammography. The proposed system can be used to enhance the mammogram and detect tumors and other related problems. There are some new image rocessing methods to extract tumor masses and clustered micro calcifications have been developed and implemented in the CAD system. The aim is to delineate regions, each of which to a certain extent uniform in some appropriate measure. A new filter called Iris Filter has been developed to detect tumor candidates. It realizes reliable detection of tumor candidates regardless of their sizes and their contrast against their background on mammograms. Previous works suggests exhaustive search technique in which we have to apply the filter in each and every pixels of the image. In this research the search space will be reduced by approximately segmenting the potential tumor areas by using suitable threshold based segmenting operation and then applying the filter only in the segmented region. In this work, apart from Iris filter, various segmentation and edge detectionalgorithms are also under consideration. Since the data storage and speed of performance issues are critical in such kind of software design, they also will be taken in to account.