Analyse multi-échelle de l'os trabéculaire vertébral à partir d'images tomodensitométriques

Abstract

The primary factor used to assess fracture risk in the clinic and diagnose osteoporosis is bone quality generally characterized by bone mineral density (BMD). However, many studies indicate that bone quality is influenced by other factors such as the microarchitecture of trabecular bone. Moreover, no tools allow convenient and simultaneous in-vivo assessment of bone density and microarchitecture, especially for the spine where penetration into deep tissues is problematic. The main objective of this thesis is to suggest a multi-scale algorithm for simulating vertebral trabecular bone microarchitecture from computed tomography (CT) images. To meet the objectives, calf, pig and human cadaveric specimens where first imaged at high and low resolution and next their microarchitecture was simulated. Mean relative errors on various morphometric parameters as well as Z-scores between those values where then quantified between high-resolution real and simulated images. The results show that the algorithm seems to reproduce well qualitatively the relatively homogeneous microarchitecture of calf and pig specimens, but poorly reproduce the relatively heterogeneous microarchitecture of human specimens. Quantitatively, the results show that the algorithm is able to reproduce several morphometric parameters. Low standard deviations of the relative errors demonstrate the high reproducibility of the simulations. However, variable sizes of Z-scores obtained for pig and human models as well as differences observed between human specimens demonstrate the high sensitivity of the algorithm. Long-term expected benefits of the proposed method are to enable more accurate clinical assessment of bone quality, to improve evaluation and planning tools for the treatment of fractures and to help diagnosis and treatment of bone pathologies such as osteoporosis.

Date	20 Dec 2013
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	Yvan Petit (Supervisor) & Erwan Gloaguen (Co-supervisor)