Rationale and Objectives

We investigated the methodologic development of computed microtomography (CMT) for monitoring the development of osteoporosis in male Sprague-Dawley rats.

Methods

Eight rats were gastrectomized and eight rats were sham operated. Femurs, tibias, and tails were prepared, and CMT scans with spatial resolutions of 5–500 μm were made. Bone diameters, bone areas, and moments of inertia were determined from the CMT scans. Optimal slice position and the need for spatial resolution and energy optimization for future in vivo applications were investigated.

Results

Gastrectomy caused dramatic changes in the bone architecture of the tibia and the femur. The main features were vacuolization of the bone and reduced amounts of compact bone. Although the outer diameters of tubular bones (femur and tibia) were largely unaffected, their inner diameters were greatly increased following gastrectomy. Relative bone area and moment of inertia were greatly reduced. The optimal photon energy was 12 keV.

Conclusion

It is possible to monitor gastrectomy-evoked changes in bone morphology at various sites in rats using CMT scanning. The changes are suggestive of osteoporosis. By optimizing the energy spectrum and spatial resolution, as well as choosing the proper slice position, it should be possible to keep absorbed doses low enough to avoid acute radiation injury in repeated in vivo measurements.

Image quality in Computerised Tomography (CT) depends strongly on the quality of the CT-projection data. These depend on sample composition and geometry, contrasting details within the sample and the equipment used, i.e. X-ray spectra, filtration, detector response and geometry. This paper focuses on the problem of selecting the optimal physical parameters to maximise the signal-to-noise in CT projection data (SNR_CT) between a contrasting detail and the surrounding sample for CT-scanners equipped with poly-energetic X-ray sources (conventional X-ray tubes) and energy-integrating detector systems (image intensifier and optical video chain). The work includes the derivation and verification ofa theoretical model for SNR_CT which can be used for predicting the optimal physical parameters for specific imaging tasks. It is shown that simplified calculations valid for mono-energetic X-ray sources and/or photon counting detectors do not correctly predict the optimal settings. This study also includes measurements of the actual X-ray source and photon transport Monte Carlo simulations of the response of the detector system.

Image quality in Computerised Tomography (CT) depends strongly on the quality of the CT-projection data, which vary with the imaged sample and the equipment used. The objective has been to find a setting of the CT-scanner Control Factors (CF) maximising the signal-to-noise ratio in CT projection data (SNR_CT) of a contrasting detail (for example a defect) and a surrounding cylindrical sample, and to present a general optimisation methodology. An optimisation case study was carried out, valid for a CT-scanner equipped with a polyenergetic X-ray source (conventional) with tungsten target and a partially energy-integrating detector system (image intensifier and optical video chain), with and without consideration qf the exposure limits associated with the microfocal X-ray source used. The CF of interest were tube potential, exposure (product of tube current and exposure time), material and thickness of the primary filter, optical aperture and attenuation equalising filter design. The settings yielding the highest values of SNR_CT were found using thick filter of high atomic number, small iris and use of an X-ray attenuation equalisation filter design. The exposure limits make the CF interdependent, yielding another optimal setting. The CF setting was also found to be independent of the contrasting detail, in the particular case study.

In this study the relationships of bone mineral density (BMD) and bone structure parameters calculated from 2D microtomography images to bone strength were investigated. Femurs from 21 male Sprague Dawley rats were subjected to dual-energy X-ray absorptiometry, computerized microtomography (CμT) and either three-point cantilever bending (femoral shaft) or two-point bending compression (femoral neck). Gastrectomy was performed on 12 animals and 9 were sham operated. From the tomograms bone structure analysis was performed using a software routine based on grey level run-length method. Correlations of BMD and bone structure parameters to mechanical parameters were investigated as were differences between the gastrectomized and the control samples. The reductions of BMD between the groups were 21 and 27% in the femoral neck and shaft, respectively. For the shaft, the correlations of BMD to all mechanical parameters were significant and BMD was a consistent predictor of bone strength for cortical bone. However, in the femoral neck where cancellous bone predominates, BMD was weakly correlated only to deflection. A significant correlation between trabecular thickness and neck bone strength was found. Hence, compared to trabecular thickness, BMD was of limited value in predicting bone strength in the femoral neck.