A subject-specific FEM to predict deep tissue mechanical stresses when supine: Development of efficient contact interfaces using Shared Topology.

Abstract

Prior studies have demonstrated Finite Element (FE) analysis is a useful tool when analysing the complex interplay of tissue and body loads which act through the human pelvis in a subject lying supine. The computational accuracy and efficiency of FE models that contain complex non-linear geometric interfaces between different anatomical and tissue regions can be compromised by superfluous node interactions and contact penetrations. This study proposes a method for the development of efficient contact definitions using shared topology. The Shared Topology Finite Element Model (FEM) resulted in a 37% reduction in solution time compared to an equivalent FEM defined with Bonded contact. At all tissue interfaces, contact penetration occurred in the Bonded FEM, with subsequent under-prediction of peak compressive strains and stresses by 1-7% compared to the Shared Topology FEM. Simulating supine lying of a 19-year-old male, the Shared Topology FEM predicted peak compressive stress in the muscle interfacing the sacrum of 29.4 kPa, and peak compressive strain of 50%. The proposed methodology can be applied for any medical imaging derived FEM where there are multiple congruent 3D geometries with negligible sliding across interfaces.