A fast methodology for generating skeletal FEM with detailed human geometric features based on CPD and RBF algorithms

Area of Science:

• Biomechanics
• Computational modeling
• Medical imaging

Background:

• Developing human body Finite Element Models (FEM) with detailed anatomical characteristics is crucial for understanding traffic accident injury mechanisms.
• Traditional FEM development is complex and time-consuming, especially the meshing process.

Purpose of the Study:

• To propose a novel, fast methodology for rapidly developing human bone FEM with detailed anatomical characteristics.
• To improve the accuracy and speed of mesh morphing in FEM generation.

Main Methods:

• A new methodology combining Coherent Point Drift (CPD) for automatic feature point generation and Radial Basis Function (RBF) for mesh morphing was developed.
• RBF-based mesh morphing generated FEM meshes from target Computed Tomography (CT) data.
• CPD algorithm's point-cloud registration rapidly generated necessary geometric feature points for mesh morphing.

Main Results:

• A 3-year-old ribcage FEM with 27,728 elements (3-5 mm mesh size) was generated in approximately 2.7 seconds.
• The average error between the generated FEM and target geometries was approximately 2.7 mm.
• The new FEM accurately represented detailed anatomical characteristics, with mesh quality comparable to the source FEM.

Conclusions:

• The proposed CPD and RBF-based methodology enables rapid and accurate development of anatomically detailed human bone FEM.
• This approach significantly reduces the time and complexity associated with traditional FEM generation.
• The generated FEMs are suitable for detailed analysis of injury mechanisms in traffic accidents.