FINITE ELEMENT MODELLING FOR THE STUDY OF PELVIC ORGAN PROLAPSE Zhuowei Chen (1), Pierre Joli (1), Zhi-Qiang Feng (1, 2) 1. LME-Evry, Université d’Evry, 91020 Evry, France; 2. School of Mechanics and Engineering, Southwest Jiaotong University, Chengdu, China Introduction Pelvic organ prolapse (POP) occurs only in women and becomes more common as women age. It refers to the loss of support to the structures contained within the bony pelvis and describes the descent of the pelvic organs into the vagina [Luft, 2006]. By some estimates, about half of parous women experience some degree of prolapse and 10 - 20% seeks medical care [Thakar et al., 2002]. However, the surgical practices remain poorly evaluated. It has been estimated that the recurrence of POP in nearly one third of women after repair surgery [Olsen et al., 1997], and that 80% of these cases recur within two years of surgery [Baden et al., 1992]. The use of precise numerical models of the female pelvic system will, in the future, provide the tools to simulate the dynamic behavior of pelvic organs. The realization of a simulator will allow surgeon to estimate the functional impact of his actions before implementation, to perform the surgery in a more controlled and reliable way. Methods In order to perform the numerical simulation, it is essential to have a good geometrical description of the pelvic cavity and to determine the material properties of organs. A geometrical rebuilding is carried out based on the magnetic resonance image (MRI) data from a single patient without genital prolapse. Among the different strain energy density functions available in the literature, Mooney-Rivlin model is chosen in the simulation of soft tissues contact [Majumder et al., 2008][Rao et al., 2010] [Noakes et al., 2008]. Its strain energy density formulation is given by 2 10 1 01 2 1 ( 3) ( 3) ( 1) W C I C I J d = - + - + - (1) where C 10 , C 01 and d are the material coefficients. 1 I and 2 I are the first and second invariants, respectively. The contact laws (Signorini and Coulomb) are formulated from an augmented Lagrangian formulation (bi-potential formulation) and computed by Uzawa or Newton techniques which lead to an iterative predictor/corrector process [Feng et al., 2003]. Results A 2D sagittal section of 3D model was chosen to be the test model. The mesh is constructed with plane quadrilateral elements. A downwards displacement of 27.9 mm [Noakes et al., 2008] was applied on the upper side of uterus. The finite element mesh of the initial step and the deformation of the last step are shown in Figure 1. The different colours mean different material properties of hyperelastic models. Figure 1: Initial and deformed meshes. Discussion The numerical results show the feasibility of dealing with large deformation contact between several pelvic organs. This work opens multiple perspectives and can propose a tool to help physicians to characterize functional impact of these organs. Many problems remain to be solved to have simulations close to reality such as having quantitative measurements of the mechanical properties of biological tissues in vivo and improving the modelling of soft tissue by considering anisotropic hyperelastic behaviours and by building 3D models with fine meshes. References Baden WF et al, JB Lippincott, 9: 235–52, 1992. Feng ZQ et al, Int J Eng Sci 41:2213–2225, 2003. Luft J, J. Nurse Practitioners, 2(3):170-177, 2006. Majumder S et al, J Biomech, 41:2834–2842, 2008. Noakes KF et al, J Biomech, 41:3060–3065, 2008. Olsen AL et al, Obstet Gynecol, 89(4):501–6, 1997. Rao GV et al, Comput Meth Biomech Biomed. Eng, 13:349–357, 2010. Thakar R et al, Br Med J, 324(7348):1256-1262, 2002. S66 Presentation 1017 - Topic 07. Biomechanics of the pelvic floor Journal of Biomechanics 45(S1) ESB2012: 18th Congress of the European Society of Biomechanics