Towards the use of meshless methods to simulate brain growth in early development

Abstract

The process by which the brain develops from a smooth surface at early weeks of gestation to a folded surface with sulci and gyri at the neonatal stage is still an unsolved question. Despite the hypotheses proposed by several theories, the lack of large data at different developmental stages and limitations in computational resources have made it difficult to develop mechanical models of brain growth to better understand which mechanisms control the cortical folding process and test different hypothesis about the most relevant factors guiding this phenomenon. An additional challenge is to work with finite-element meshes (FEM) in this application, due to the large deformations undergone in brain development that can generate overlapping elements, which hamper obtaining the correct solutions and prevent the simulations to converge. Thus, meshless modelling techniques could give a solution to that issue since they are based on particles, not requiring node connectivity and being robust against complex mechanisms such as brain folding. The main goal of this TFG is setting the first steps that could allow modelling cortical folding through meshless methods. To do so, a Smoothed Particle Hydrodynamics (SPH) environment built for modelling cardiac mechanical motion was adapted in order to incorporate the mechanical equations that describe the gyrification process. Then, the results were assessed by making a comparison with the current FEM-based implementations in simplified synthetic scenarios.