Measures of complexity and processing of vertebral CT-images

The authors propose an algorithm which processes computed tomography images of vertebral bodies and quantifies their structure. Vertebral bodies are segmented from the connective and soft tissue background, and then the image of the entire vertebrae are split into the cortical and trabecular bone. At the next stage, several criteria, based on nonlinear dynamics, complexity measures and symbolic dynamics, are applied. These measures can contribute significantly to the early diagnostics of changes in bone structure, which are specific for osteoporosis and other bone diseases