The steps followed in this software selleck chemicals Enzastaurin were as follows.(i) Volume Reconstruction ��The 3D volume of the breast was reconstructed from the all segmented tissues from the segmented DICOM image imported by Simpleware.(ii) Smoothness ��The smoothing recursive Gaussian filter with sigma values between 0.4 and 0.7 in the X, Y, and Z directions was used in order to smooth the surfaces of the three tissues, the island removal filter was applied to classify free pixels that still could keep in the breast, and the cavity fill filter was applied to fill possible hollows.(iii) Mesh Generation ��An algorithm that allows surface adaptation was used in order to achieve the best mesh optimization, reducing the number of elements.
In the case of considering the skin as a small surface of constant thickness, the real skin was removed and replaced by a 2D membrane of uniform thickness which covered all the breast.(iv) *.ans File Generation ��A file with all the information about the meshes of the three tissues was generated in Simpleware. This file can be read by ANSYS. ANSYS loads the meshes and allows the deformation to be simulated.Figure 5 shows the meshes of the glandular and skin tissues for the two cases under study: considering real skin on the left and considering the skin as a 2D membrane on the right. The elements chosen to construct the 3D meshes of the three tissues were SOLID186, except when the skin was assumed as a 2D shell. In this case, the chosen element was the shell element SHELL181 with thickness of 1mm [10, 25].
Figure 5Meshes of the glandular and skin tissues for the two cases under study: (a), considering real skin and, (b), considering the skin as a shell.To simulate the biomechanical behavior of the breast tissues under compression, the hyperelastic model used for the dense and glandular tissues by del Palomar et al. in [10] was also used in this paper. Regarding the skin, the also hyperelastic model proposed by Hendriks et al. for the human skin in [26] was used. For the three cases, the model was a neo-Hookean model, for which the form of the strain energy potential, WNH, is defined by (3):WNH=C1(I��1?3)+1d(J?1)2,(3)where C1 = ��0/2 and ��0 is the initial shear modulus of the material, I��1 is the first deviatoric strain invariant, d is a material incompressibility parameter that is related to the initial bulk modulus K0 = 2/d, and J is the determinant of the elastic deformation gradient. The elastic constants used for the three tissues were the constants obtained by these authors in their respective works: C1 = 3kPa for the fat tissue, C1 = 12kPa for the glandular tissue, and C1 = 50kPa for the Batimastat skin. The values of d were obtained using the approximation to incompressible materials.