Contour-based surface modeling and analysis of microtomographic trabecular bone images

Bone quality is an important factor of bone strength. Osteoporosis is a disease characterized by longitudinal changes in 3-D trabecular bone structure, such as decreased trabecular thickness. With the purpose of identifying parameters indicate the quality of bone micro-architecture, a strategy are developed, using High Resolution Micro-Computed Tomography (μCT) images. The main idea is the use of the morphological operator and the watershed algorithm in order to separate at each slice the bone phase from the marrow phase. As well, we make use of basic morphological transforms for eliminating artifacts and aliasing from obtained contours. Three dimensional reconstructions of samples were created by stacking all the 2D obtained boundary top of each other. The zero iso-surface embedded in the computed volume in order to localize vertices from successive slices. A mesh that approximates the zero isosurface is then extracted from the resulting volume. Quantization transforms the trabecular bone structure in a set of numerical parameters that can directly gives a valid and reliable diagnosis histologist.