[Material selective imaging and density measurement with the dual energy method. III. Determination of bone mineral of the spine with CT].

Quantitative computed tomography for measuring vertebral bone mineral content offers high sensitivity and reproducibility. However, several systematic errors of the method have to be considered, such as the influence of fat present in the spongiosa in varying concentrations as well as beam hardening effects and different calibration methods. The unknown fat content of the spongiosa of the vertebral bodies can be responsible for errors up to 30%. Beam hardening effects result in comparatively minor errors which can moreover be easily corrected via reference phantoms. The calibration method, on the other hand, results in potentially major systematic deviations due to the necessarily arbitrary determination of the reference points. These deviations must be taken into account especially in cross-comparisons of measured data. Dual energy methods enable a highly selective determination of the density of calcified tissue and soft tissue and decisively reduces the systematic errors mentioned above. The order of magnitude of the individual errors for different CT techniques was assessed by means of phantom studies and clinical studies.