3D Image Analysis and Pattern Recognition for Modeling Trabecular Bone Microstructures

Trabecular bone is composed of a complex 3D microstructure which can be acquired in great detail by high resolution imaging techniques such micro computed tomography (µCT) or micro magnetic resonance imaging (µMRI). In this project, a distinctly new approach to handle microstructural modeling of trabecular bone is proposed.

Trabecular bone is composed of a complex 3D microstructure which can be acquired in great detail by high resolution imaging techniques such micro computed tomography (µCT) or micro magnetic resonance imaging (µMRI). In this project, a distinctly new approach to handle microstructural modeling of trabecular bone is proposed. We are developing image analysis procedures on detailed µCT images of trabecular bone for bone tissue segmentation, bone surface compression, bone volume skeletonization and feature analysis for rods/plates such that a reduced finite element microstructural model can be produced using beam/plate structural elements alone. In this approach, bone volume fraction, trabecular orientation and connectivity will be maintained while we expect that a 200-1000 fold reduction of the model size such that the nonlinear behavior of a large bone sample (even whole bone organ) can be analyzed. In the current study, the accuracy of the developed algorithms was validated in idealized trabecular rod and rod-plate models.

Related Publications

  1. Wei, L., Sajda, P., Laine, A. F. and Guo, X. E. A Novel Approach to Model Trabecular Bone Using Topological Image Analysis, 49th Orthopaedic Research Society Annual Meeting, 2003.