Osteoporosis is characterized by low bone mass and deterioration in microarchitecture of trabecular bone. Current methods for diagnosis of osteoporosis take little into account regarding the effects of individual trabecular rods and plates and are limited by resolution, cost, time, and safety issues (e.g. radiation exposure). In this study, we propose a novel probabilistic approach to analyze the architecture of trabecular bone and its correlation with the total bone volume fraction of the tissue.
Briefly, bone samples from ten human donors, five high (N=5) and five low (N=5) bone volume fractions (BVF) were micro-CT or serial milled imaged and spatially decomposed into individual plates and rods using an individual trabecular segmentation (ITS)-based morphological technique. The architectural parameters (i.e. spatial distribution of trabecular joint locations, orientation of trabecular plates and rods, and plate thickness and rod diameter) were analyzed visually using 1 and 2D histograms and mathematically by best-fitting the probability density functions (PDFs) of the parameters.
The results of this study indicate that the spatial distribution of joints, the orientation distribution of plates and rods, and rod diameter are independent of BVF, implying that these features of trabecular architecture are preserved even with a loss in bone mass. In contrast, the number of plates and rods as well as the thickness of plates exhibit high correlations with BVF (r= 0.85, 0.92, 0.96, respectively), thus suggesting the major contribution to bone mass loss by these parameters. In addition, trabecular plate thickness was found to be dependent on BVF. The information obtained in this study will help in prediction of bone fracture risks and can be used in the future development of representative models of trabecular bone.