Osteocytes, the most numerous bone cells, are thought to be actively involved in the bone modeling and remodeling processes. The morphology of osteocyte is hypothesized to adapt according to the physiological mechanical loading. Three-dimensional micro-CT has recently been used to study osteocyte lacunae. In this work, we proposed a computationally efficient and validated automated image analysis method to quantify the 3D shape descriptors of osteocyte lacunae and their distribution in human femurs. Thirteen samples were imaged using Synchrotron Radiation (SR) micro-CT at ID19 of the ESRF with 1.4 μm isotropic voxel resolution. With a field of view of about 2.9 × 2.9 × 1.4 mm³, the 3D images include several tens of thousands of osteocyte lacunae. We designed an automated quantification method to segment and extract 3D cell descriptors from osteocyte lacunae. An image moment-based approach was used to calculate the volume, length, width, height and anisotropy of each osteocyte lacuna. We employed a fast algorithm to further efficiently calculate the surface area, the Euler number and the structure model index (SMI) of each lacuna. We also introduced the 3D lacunar density map to directly visualize the lacunar density variation over a large field of view. We reported the lacunar morphometric properties and distributions as well as cortical bone histomorphometric indices on the 13 bone samples.
The mean volume and surface were found to be 409.5 ± 149.7 μm3 and 336.2 ± 94.5 μm². The average dimensions were of 18.9 ± 4.9 μm in length, 9.2 ± 2.1 μm in width and 4.8 ± 1.1 μm in depth. We found lacunar number density and six osteocyte lacunar descriptors, three axis lengths, two anisotropy ratios and SMI, that are significantly correlated to bone porosity at a same local region.
The proposed method allowed an automatic and efficient direct 3D analysis of a large population of bone cells and is expected to provide reliable biological information for better understanding the bone quality and diseases at cellular level.