Date on Master's Thesis/Doctoral Dissertation


Document Type

Doctoral Dissertation

Degree Name

Ph. D.


Mechanical Engineering

Committee Chair

Voor, Michael John

Author's Keywords

Trabecular bone; Bone; Micro-CT scanning




A combination of three techniques – high resolution micro computed tomography (micro CT) scanning, Archimedes-based volume fraction measurement and serial sectioning or milling – were used to determine the volume fraction, trabecular thickness, trabecular separation, trabecular number and micro finite element analysis combined with mechanical testing was used to determine the apparent stiffness and tissue modulus to quantify bone quality in rabbit distal femur trabecular bone. The objectives of this dissertation were two-fold. First, to develop the capabilities of micro CT scanning and micro CT image segmentation based on a slice-by-slice global thresholding technique to investigate trabecular microstructural changes in vivo and in vitro; and second, to develop the capability of translating micro CT scans into three dimensional finite element models based on direct voxel conversion technique. These results were validated within the in vivo and in vitro scans at the same time, and validated with the Archimedes-based volume fraction measurements and serial sectioning or milling experiments. The micro FE models were executed as linear analyses and the same bone cubes of the models were mechanically tested (compressive testing) to determine the correct tissue modulus of the bone specimens. The apparent stiffness of these micro FE models was recalculated using the average tissue modulus. A total of six six-month-old New Zealand white rabbits were utilized in this study. Three rabbits were scanned twice in vivo seven days apart (T1 and T7) and three rabbits were only scanned once in vivo. All of the femurs were scanned in vitro. All micro CT images were obtained at 28 um (in vivo) or 14 um (in vitro) nominal resolutions. Specimens from six left and right rabbit distal femurs (medial and lateral) were measured based on Archimedes' principle and serial milling. The volume fraction for lateral condyles between two in vivo scans T1 (0.401+0.015) and T7 (0.397+0.021), between in vitro micro CT (0.352+0.035) and Archimedes (0.365+0.031) and between in vitro micro (0.352+0.035) and serial milling (0.369+0.031) were not significantly different. The medial condyles were also not significantly different: T1 (0.513+0.010), T7 (0.515+0.011), in vitro micro CT (0.454+0.049), Archimedes (0.460+0.060) and serial milling (0.467+0.505). Specimens from another six left and right distal femurs (medial and lateral) were mechanically tested along the anterior-posterior directions. The tissue modulus of each specimen was determined by making the calculated apparent stiffness values from FEA to be equal to mechanical compressive testing (MTS). Based on a new constant tissue modulus, the recalculated FEA apparent stiffness (1.77E9+6.45E8) and MTS apparent stiffness (1.76E9+7.37E8) were linearly correlated (r-value = 0.8721). These findings suggest that the capabilities of slice-by-slice global thresholding and direct voxel conversion are sensitive, reliable and consistent for the study of trabecular bone microstructural changes in vivo utilizing high resolution (< 28 um) micro CT scanning and micro FEA.