In this chapter we have shown the benefits of considering the internal breast density of the mammograms. The comparison between the results obtained using this information clearly outperforms the ones obtained without using it.
Moreover, the robustness of our mass detection approach has been reported in the sense that the algorithm is able to find masses in a mammogram independently of the database used for training the system. However results show that the false positive approach is more sensitive to the database used. We consider that this is due to the nature of both algorithms: the template matching is based on the adjustment of the contours of the template and the image, while the false reduction approach is mostly based on region information (grey-level values) which are highly dependent on the database used.