Multiresolution enhanced techniques in digitized mammography

In this paper we study the application of selected multiresolution techniques to improve the contrast of on regions of interest in digitized mammography. Enhanced techniques to increase the contrast among background and lesions are necessary to improve the performance in the abnormalities detection process. Lesions enhancement is performed applying two techniques. In the first technique, the ROIs contrast is improved by applying an one-dimensional histogram multiresolution technique based in Haar transformation. in the second technique the ROIs contrast is improved applying a two-dimensional multiresolution enhanced technique based in a wavelet transformation filter bank. Both techniques are evaluated by a well-known criteria of seperability measure (DSM) and results are discussed.