A great number certainly of groups have proposed different biomechanical models of the breast for different applications. Some groups have used homogeneous models [3�C7]. However, there are other groups that, in searching for more realistic and accurate models, have proposed heterogeneous models, that is, models that take into account the three tissues of the breast. For example, Ruiter et al. [8] tested different heterogeneous model combinations (from linear to exponential) to register MRI with X-ray mammograms, Kellner et al. [9] used a linear elastic model for each tissue to simulate the mechanical compression of the breast, del Palomar et al. [10] used a neo-Hookean model for the fat and glandular tissues and a Polynomial model for the skin tissue, aimed to study the effect of gravity for surgical planning, and Tanner et al.

[11] applied a heterogeneous model consisting of linear elastic and exponential material models also for the register between MRI and X-ray mammograms. Nevertheless, although these models have considered the real fat and glandular tissues obtained, for example, from an MRI or CT, none of them have considered the real skin, but the skin has been approximated as a membrane of constant thickness which covers all the breast.To the authors’ knowledge, there are no studies that compare significant differences when an approximate skin or a more accurate one is used in the simulation of the mammographic compression. While the distribution of fat and dense tissues have been many times represented using the finite element methods, modeling the influence of the skin in the breast biomechanical models still requires further investigation.

The simplest options have involved modeling the skin layer as additional 3D elements that surround the internal tissue elements or coupling 2D membrane (skin) elements to 3D fat/dense elements. However, the skin obtained with these approximations is a constant surface which does not take into account issues like widened skin due to suspicious masses near skin area [12, 13] or a differentiated nipple region. Therefore, there is a need to develop a segmentation method for the skin able to remove the real skin (i.e., the external tissue surrounding the breast) from the rest of the breast. This method is aimed to let a correct segmentation of the breast in order to construct an accurate biomechanical model that can be used in the different simulations of the different imaging modalities.In this paper, a novel method Drug_discovery for a more accurate skin segmentation is presented.