Bone is a hierarchically organized, multiscale composite with remarkable abilities to repair, adapt and maintain itself. In the bone cortex, the basic structural and functional unit is the osteon, a cylindrical layered structure surrounding a vessel canal (diameter ≈ 200-300 μm, in human bone). The osteocytes are the most abundant and longest living bone cells, embedded in the mineralized bone matrix and regularly distributed within the osteons. They are interconnected with each other and with the vessel canals through dendrites, located in slender canals called canaliculi. The osteocyte lacunae, cavities in which the cells are located, together with the canaliculi form a communication network throughout the bone matrix, permitting transport of nutrients, waste and signals. These cells were considered passive for a long time, but in the last decades it has become increasingly clear their role as mechanosensory cells and orchestrators of bone remodeling, and hence their responsibility for bone quality. Today great interest is given to the osteocytes and the lacuno-canalicular network.

Despite recent advances in imaging techniques, none of the available methods can provide an adequate 3D assessment of the lacuno-canalicular network. This is due to its situation in the hard mineralized matrix, combined with its characteristic 3D complexity, the nanoscale size of canaliculi (~ 300-900 nm in diameter in human bone) and the necessity to assess it in a relatively large volume of tissue, in order to achieve a statistically significant evaluation. In this context, many questions about the 3D architecture of the osteocyte network, and its relation with age, disease or mechanical loading, are unanswered.

The aims of this thesis were to achieve three-dimensional (3D) imaging of the osteocyte lacunocanalicular network with synchrotron radiation X-ray computed tomography (SR-CT) and to develop tools for 3D visualization and segmentation of this cell network, leading towards automatic quantification and analysis of this structure.

We propose to use parallel beam synchrotron X-ray computed tomography to image in 3D the lacuno-canalicular network in bone. This technique can provide 3D data on both the morphology of the cell network and the composition of the bone matrix. With a pixel size set at 280 nm and a 2D detector composed of 2048×2048 elements, the imaged volume corresponds to ~ 5743 μm³. Compared to the other 3D imaging methods with comparable spatial resolution, this enables imaging of tissue covering a number of cell lacunae three orders of magnitude greater in a single image, in a simpler and faster way. This makes possible the study of sets of specimens in order to reach biomedical conclusions. The main challenges to surpass in order to attain imaging of this structure with SR-CT are related to the very high spatial resolution required to resolve the canaliculi (minimum pixel size reachable is 280 nm), combined with difficulties related to radiation exposure. We could establish valid imaging setups and protocols, based on two insertion devices.

Furthermore, we propose the use of a new synchrotron X-ray tomography technique, dubbed magnified holotomography, to image the ultrastructure of bone tissue. By using a divergent beam, this method can reach a higher resolution (nominal pixel size 60 nm, field of view ~ 1203 μm³). The tomographic reconstruction is no longer based on the angular X-ray attenuation maps, but on phase maps, obtained after the application of a suitable phase retrieval algorithm. The phase is retrieved from projections recorded at four different distances between the sample and the detector, for each rotation angle. The reconstructed image corresponds to the 3D distribution of the complex refractive index of the specimen, which is related to the 3D electron density. This technique permits assessment of the cell network with higher accuracy and it enables the three-dimensional organization of collagen fibres organization in the bone matrix to be visualized for the first time.

In order to obtain quantitative parameters on the geometry of the cell network, this has to be separated from the bone matrix. The segmentation needs to be automatic, given that each image contains about 3 000 cell lacunae with more than 100 000 canaliculi. Due to the limitations in spatial resolution, canaliculi appear as 3D tube-like structures measuring only one to three voxels in diameter. This, combined with the noise and the low contrast of the images, and the large size of each image (8 GB), makes the segmentation a difficult task.

We propose an image enhancement method, based on a 3D line filter combined with bilateral filtering, which reintroduces in the filter map the grey level information from the original image. This enables improvement in canaliculi detection, reduction of the background noise and cell lacunae preservation. For the image segmentation we developed a method based on variational region growing. We propose two expressions for energy functionals to minimize in order to detect the desired structure, based on the 3D line filter map and the original image. Future work will focus on improving the segmentation results and the development of automatic quantification techniques.

Preliminary quantitative results are extracted based on a connected components analysis and a few observations related to the bone cell network and its relation with the bone matrix are presented.