MRI bone imaging and analysis.

• Main Author: Chua, Nanzhi.
• Other Authors: Poh Chueh Loo
• Format: Final Year Project
• Language: English
• Published: 2013
• Subjects: DRNTU::Engineering::Bioengineering
• Online Access: http://hdl.handle.net/10356/53685
• Institution: Nanyang Technological University

The aim of this study is to investigate the capability of magnetic resonance imaging (MRI) as a diagnostic tool for osteoporosis and also to evaluate the potential of improving trabecular bone characterization with image processing techniques. Twenty-one cylindrical trabecular bones were obtained from goat’s femur bone and scanned with micro-computed tomography (μCT) and 3 Tesla MRI machine. In this study, result from μCT will be the standard of reference. Super resolution (SR) volume reconstruction, which is an image processing technique, was applied on MR images. Trabecular bones were segmented from marrow and empty space for MR images by K-means clustering. Derivation of the four histomorphometric parameters namely bone volume fraction (BVF), trabecular thickness (Tb.Th), trabecular number (Tb.N) and trabecular spacing (Tb.Sp) were done thereafter. Bone mineral density (BMD) of the specimens was then acquired by dual energy X-ray absorptiometry (DXA). Linear regression analysis was conducted to analysis the correlation of the parameters. MRI correlation were R2= 0.5747 for BVF, R2= 0.0746 for Tb.Th, R2= 0. 2414 for Tb.N and R2= 0.1785 for Tb.Sp. The correlation for MRI after going through SR volume reconstruction were R2= 0.8622 for BVF, R2= 0.0121 for Tb.Th, R2= 0. 0566 for Tb.N and R2= 0. 353 for Tb.Sp. A significant increase in the correlation for BVF after SR volume reconstruction was observed. Also, an increase in correlation of BMD after SR volume reconstruction were observed with R2= 0.5336 for normal MRI and R2= 0.6528 for MRI after SR volume reconstruction. The result of this study proved that SR volume reconstruction is able to improve the correlation of MRI derived parameters with μCT derived parameters.