Real-time simulation of mechanical properties on virtual reality: A methodology to improve geometric segmentation, mathematical modeling and characterization of soft tissues

Abstract

In this research project, it is presented a methodology to achieve the development of technological tools and material knowledge that support the real-time simulation of soft tissues for virtual reality. Particularly, this work is focused on two main areas that were identified in previous work as opportunities for improvement: Geometrical and Material Modeling. These areas are key to develop not only medical training processes, but also other research projects that involve soft tissue and composite materials characterization, design and development of biomedical devices, and augmented reality tools, among others. As one of the goals, it is proposed to create virtual tools that allow the interaction, processing, and segmentation of medical images in a semi-automatic way. This was detected by questioning how to increase the applicability of the simulation framework to other anatomical geometries and simplify the creation of new and customized medical cases based on their own set of images. The solution proposed is to provide the user with access to an interactive learning experience based on 3D rendering of medical images. This will not only allow visualization of medical cases but also have a relatively quick and simple process to get anatomically realistic 3D geometries for simulation, design of new products or 3D printing of models. In this path, it was developed the module VISUALIX, which is able to provide said interaction, and the results are presented in chapter 4. Also, a proposal for fractal structure analysis was done using microtomography images, creating FractalCells module with the implemented tools. For material modeling and characterization of soft tissue, a new hybrid formulation is proposed by questioning how a simple technique like Spring-Mass Model (SMM) can describe the soft tissue mechanical properties, if it is based on linear elasticity theory and therefore it can only be used to predict small deformations (<10%). The solution proposed is based on the application of a constitutive model able to describe the mechanical behavior of soft tissue, as well as other composite materials. For this purpose, it is created a hybrid construction of a Strain Energy Density Function (SEDF) used to find an energy equivalence with a variable stiffness SMM. This formulation was named Equivalent Energy Spring Model (EESM) and is Abstract ii presented in chapter 5. It is able to characterize soft tissue properties of non-linearity, anisotropy, and Mullin’s effect, to predict its response at large deformation (>10%). Finally, and in order to validate the proposed model, an experimental phase was defined to perform uniaxial and biaxial cyclical tensile tests with porcine tissue samples in order to have experimental data for material characterization using EESM. This experimental phase is described in chapter 6. The results for the characterization of porcine liver and abdominal wall tissues, as well as the predictions of the EESM formulation are presented in chapter 7, including an evaluation of its accuracy and its capacity to perform in real-time.