Physics in Medicine and Biology · 1994 · 246 citations · 26 references
EngineeringBioengineeringBiomedical EngineeringBiomedical ImagingDiagnostic ImagingDiagnostic RadiologyBone MechanicsComputed TomographyMedical ImagingThree-dimensional ReconstructionVolume Segmentation3D Medical Image AnalysisThree-dimensional Bone ImagingThree-dimensional Medical ImagingTrabecular ArchitectureTrabecular Bone AnalysisTrabecular Bone ArchitectureVolumetric Bone AnalysisVolumetric Data Acquisition
Three-dimensional structure is one of the main factors influencing the mechanical behaviour of cancellous bone. To analyse the trabecular bone structure non-destructively we used a peripheral QCT system and applied a special thin-slice technique to create high-resolution volumetric data sets serving as a basis for something we would like to call non-invasive bone biopsy. In order to obtain binary data sets, the mineralized bone in the CT volume was separated from bone marrow and muscle tissue with the help of a sophisticated three-dimensional segmentation algorithm based on the analysis of directional derivatives, which are computed from a locally approximated fit function of the original CT volume. Binary volumes including either a solid representation of trabecular plates and rods or a topological representation of the cancellous bone architecture were acquired. Such volumes can be processed non-destructively and, even more important, repetitively. By using a surface reconstruction algorithm based on interpolating triangulation it was possible to visualize the three-dimensional surface of the trabecular bone structure. The results showed that surface representation and visualization in combination with a multiple thin-slice measuring technique are valuable tools in studying three-dimensional bone architecture. In the future, the non-invasive bone biopsies will be evaluated by means of three-dimensional mechanical analysis incorporating finite element modelling and direct morphological investigations of the cancellous bone architecture for a better prediction of bone strength as an index for fracture risk or osteoporosis.
26
Marching cubes: A high resolution 3D surface construction algorithm
William E. Lorensen, H. E. Cline · 1987 · 10.2K citations · Full text
Computing, Radiology, Visualization +17
A. M. Parfitt, Catharina H. E. Mathews, A. R. Villanueva et al. · Journal of Clinical Investigation · 1983 · 1.5K citations · Full text
A survey on image segmentation
K. S. Fu, Jack K. Mui · Pattern Recognition · 1981 · 1.2K citations