Multiresolution analysis has been useful in so many applications from image compression to image de-noising and
classification. Several methods using wavelet have been proposed for feature extraction in mammograms. Liu et al. proved that multiresolution analysis of mammograms improve the effectiveness of the diagnosis system based on wavelets coefficients. In their mammogram analysis study, they use a set of statistical features with binary tree classifier
in their diagnosis system to detect the spiculated mass. The achieved successful rate was 84.2%. Mousa et al. proposed system based on wavelet analysis and used the Adaptive Neuro-Fuzzy Inference System (ANFIS) for building the classifier to distinguish between mass and microcalcification, the maximum classification rate obtained was 85.4%. Rashed et al. studied the multiresolution analysis of digital mammogram using wavelet transform. They used Euclidean distance to classify between micro-calcification clusters, spiculated mass, circumscribed mass, ill-defined mass and normal mammogram. The maximum classification rate achieved was 87.06%. Ferreira and Borges [9] proposed system to classify the mammogram images by transforming the images into wavelet bases and then using a sets of coefficients from first level of decomposition as the feature vector toward separating micro-calcification clusters, spiculated mass, circumscribed mass, and normal classes of image. The maximum classification rate achieved was 94.85%. Curvelet was developed by Candes and Donoho [10], for providing efficient representation of smooth objects with discontinuities along curves. Detecting and enhancing the boundaries between different structures is very important in image processing, especially in medical imaging. Some studies using curvelet transform in image processing have been done. Ali et al. presented a curvelet approach for the fusion of Magnetic Resonance (MR) and Computed Tomography (CT) images. They found that curvelet transform achieved good results in their fusion. Bind and Tahan presented a method for object detection of speckle image based on curvelet transform. They constructed a segmentation method which provides a sparse expansion for typical images having smooth contours. Murtagh and Stark used second, third and fourth order moment of Multiresolution transform (wavelet and curvelet) coefficients as features, and K-nearest neighbors supervised classifier for image classification process.