Voronoi-based analysis of bone cell network from synchrotron radiation micro-CT images

P. Dong (1) , Sébastien Valette (2) , M. A. Zuluaga (3) , J. Kazakia G. (4) , F. Peyrin (1, 5)

1 Imagerie Tomographique et Radiothérapie
2 Images et Modèles
3 CMIC - Centre for Medical Image Computing
4 Department of Radiology and Biomedical Imaging [San Francisco]
5 ESRF - European Synchrotron Radiation Facility

With the development of the novel micro and nano-CT systems, bone cell analysis has gone beyond the limitation of the conventional 2D analysis. In particular synchrotron radiation micro-CT is well suited to image in 3D the lacuno-canalicular network (LCN) in bone tissue. This network is made of osteocyte lacunae connected by small channels called canaliculi. Due to the lack of quantitative data on this network, we propose here an automated method to extract geodesic Voronoi-based parameters to characterize the regions of influence of canaliculi. To this aim, after labeling each lacuna from the segmented LCN image, we generated geodesic Voronoi tessellations on each lacunar surface. Our proposed method was successfully applied to three SR micro-CT images of women tibial cortical samples. We believe that this method can serve to extract new information on the 3D morphometry of the LCN in more datasets.

Domaines: Traitement du signal et de l'image [eess.SP]; Imagerie

P. Dong, Sébastien Valette, M. A. Zuluaga, J. Kazakia G., F. Peyrin. Voronoi-based analysis of bone cell network from synchrotron radiation micro-CT images. Biomedical Imaging (ISBI), 2015 IEEE 12th International Symposium on, Apr 2015, New-York, United States. pp.1522 - 1525, ⟨10.1109/ISBI.2015.7164167⟩. ⟨hal-01272791⟩