Curvelet transform is used as a multi-scale level decomposition to represent mammogram images. The calculated texture features are used as feature vector of the corresponding mammogram. Table III shows the successful classification rate of mammogram images with the overall classification accuracy based on 5-fold cross validation. The average rate for each fold is calculated then the average for 5-folds is calculated. It shows that, the successful classification rate of mammogram images for normal and abnormal using nearest neighbor classifier reached to 99.03% in fold 4, while the average classification rate achieved for all folds is 97.03%. For the second step, the classification rates of the abnormalities of 5-fold cross validation are listed in Table IV. The average rate for each fold is calculated then the
average for 5-folds is calculated is calculated. Table IV shows that, the average successful classification rate for both classes reaches to 95.10% in fold 4, while the average classification rate achieved for all folds is 91.68%. It can be concluded that the results show that the proposed method is able to find an appropriate feature set that lead to significant improvement in classification accuracy. We believe that the high successful classification rate achieved is a result of using curvelet transform. This supports the claim that curvelet transform provide stable, efficient and near-optimal representation of smooth objects having discontinuities along smooth curves.