Classification of Breast Masses in Mammograms using Support Vector Machine
G. Suganthi, J. Sutha · 2012
A Multi view CADx System for the mammography images is implemented. Two types of systems are widely used in mammography. They are Computer-aided detection (CADe) and Computer-aided diagnosis (CADx). The different views of mammography images MLO (Mediolateral Oblique) and CC (Crani-caudal) are assessed. Segmentation will be implemented for the images obtained from the two views for extracting the mass contour. A set of features related to the geometry of the boundary and the structure inside it will be computed for both of the images. An optimal subset of similar features will be extracted. Using the ranked features extracted the classification will be implemented using SVM. A Monte – Carlo method owing to the iterative and complex structure of the algorithms is used. The validation of the results is based on confidence intervals for given coverage probabilities and performance metrics.