Bone fracture prediction using computed tomography based rigidity analysis and the finite element method

In this study two imaging modalities are proposed as a basis for fracture analysis in bone; Computed Tomography (CT) and Positron Emission Tomography (PET).;In the first part of this study, finite element analysis (FEA); CT based structural rigidity analysis (CTRA) and mechanical testing are performed to assess the immediate fracture risk of rat tibia with simulated lytic defects at different locations.;Twenty rat tibia were randomly assigned to four equal groups (n=5). Three of the groups included a simulated defect at various locations: anterior bone surface (Group 1), posterior bone surface (Group 2), and through bone defect (Group 3). The fourth group was a control group with no defect (Group 4). Micro computed tomography was used to assess bone structural rigidity properties and to provide 3D model data for generation of the finite element models for each specimen.;Compressive failure load was predicted using CT derived rigidity parameters (FCTRA) and was correlated to failure load recorded in mechanical testing (R2=0.96). The relationships between mechanical testing failure load and the axial rigidity (R2=0.61), bending rigidity (R2=0.71) and FEA calculated failure load (R2=0.75) were also correlated. CTRA stress, calculated adjacent to the defect, were also shown to be well correlated with yield stresses calculated using the average density at the weakest cross section (R2=0.72). No statistically significant relationship between apparent density and mechanical testing failure load was found (P=0.37).;In the second part of this study, Positron Emission Tomography (PET) and Computed Tomography (CT) imaging modalities were utilized to study the implementation of a bone remodeling rule for the study of future fracture risk.;Eight rats were inoculated with MDA-MB-231 human breast cancer cells at T0 to induce osteolytic lesions that simulate skeletal metastasis. Fluorine 18 (18F) and fluoro-deoxy-glucose (FDG) PET and CT imaging were carried out weekly to correlate changes in local bone mineral density observed using CT, with radionuclide tracer uptake observed in 18F) and FDG PET. Univariate relationships correlating pixel level density to standardized uptake values (SUV's) were obtained (R2=0.4--0.72). These relationships can be applied in assigning material properties in finite element models studying future fracture risk. Furthermore, a tumor induced bone remodeling rule could be developed to allow determination of future fracture risk associated with a baseline CT scan taken at time T0.;In summary, the results of this study indicate that CTRA analysis of bone strength correlates well with both FEA results and those obtained from mechanical testing. In addition there exist a good correlation between structural rigidity parameters and experimental failure loads. In contrast, there was no correlation between average bone density and failure load. Furthermore, the positron emission tomography (PET) imaging modality offers a promising method of studying future fracture risk using the finite element method.