Automatic segmentation of cortical and trabecular components of bone specimens acquired by pQCT

Abstract

Peripheral Quantitative Computerized Tomography (pQCT) allows the acquisition of bone specimens with a spatial resolution adequate to visualize the 3D structure of the bone cortex and the trabecular network. At present, pQCT scanners are equipped with image processing software that limits the bone analysis in two dimensions and requires strong user interaction. In this work, a method is proposed to automatically segment, in 3D, cortical and trabecular components of bone specimens acquired by pQCT, in order to facilitate and enhance the quantitative evaluation of densitometric properties of the bone.