Automation of Bone Tissue Histology

In abstract methods of automation of histology, bone structure is considered. Possible inputs are snapshots from a microscope or computed tomography slices. An algorithm is proposed that differentiates objects according to their color (or grayscale) and recover morphology topology. An algorithm to separate morphological objects by their dimensions and color parameters was built. Measured parameters were bone surface, bone area, porosity, cortical thickness, canal number, canal area, and etc. Additionally, we measured the anisotropy properties of the bone tissue: distribution of porosity direction and degree of porosity elongation. A bone example was scanned by computed tomography. All data were measured by the proposed method and the results presented. An example algorithm of work on computed tomography data is shown in this work.