Development of a Framework Employing Open-Source Software for Finite Elements in Computational Biomechanics

Abstract

This work represents an investigation and demonstration of a strategy, using a pipeline of open source software for computational biomechanics. The purpose is to extract 3D models from MRI data and perform Finite Element Analysis and ensure faithfulness in open simulation. The methodology considers four robust open-source workbenches including 3D Slicer (for segmentation), Blender (for cleaning and fixing STL models and setting up physical boundary conditions), Salome-Meca (for Finite Element Analysis and Post-Processing) and Paravis (for Post Processing and Visualization). The model for this study consisted of a vertebra and an inter-vertebral disk from human spine. Raw model was extracted from 3D slicer. Then it was pre-processed in Blender and finally finite element analysis was performed in Salome-Meca. While conceiving the initiative of free pipeline for this particular study in computational biomechanics, it was assured to guarantee the full control of the implemented functions. Not only this, the parameterization and interchangeability between the workbenches was also entrusted. Fixing serious mesh problems like singularity, free and intersecting elements is special contribution of this study using open source workbench. Generation of contact surfaces and categorization in groups, pivotal for assemblage later in Code_Aster, with a Blender script was also implemented which is very similar to how they are done in licensed workbenches for FEM. Another major highlight of this work is a novel strategy to separate a vertebra from the spinal column using 3D slicer. The results from Code_Aster were cross validated with with commercial computer solvers. In conclusion, apart from free software benefits, it is a recommended solution for all professionals in the field of computational biomechanics and relative researches where mathematical modeling is an emerging field and ability to customize the software shouldn’t be limited.