Abstract:
This paper presents a new method for automating the generation of volumetric grids, or meshes, which are essential for the finite element modeling of human hand joints. The research is motivated by the need to make sophisticated biomechanical analysis tools faster and more accessible for clinical applications, such as diagnosis and treatment planning, by reducing the time-intensive manual labor typically required for model creation. The proposed technique relies on a library of pre-defined meshing "templates" that correspond to different anatomical structures.
To create a patient-specific model, an appropriate template is selected and then automatically deformed to fit the geometry of a specific bone, which has been captured from medical images like an MRI. The gridding algorithm for a long bone involves first defining a central "spine" along the bone's axis, then using a series of perpendicular and radial planes to guide the projection of a computational mesh onto the bone's surface. The internal nodes are then optimized to ensure a high-quality grid suitable for finite element analysis. The authors successfully demonstrate the validity of this approach by generating high-quality meshes for all the finger bones in a hand data set. This work represents a critical first step in automating the creation of complex, patient-specific biomechanical models for any tissue in the human body.
