Phantom studies in osteoporosis.

Differences in measurement results of bone densitometry are an obvious disadvantage of this method. The differences are mainly due to the calibration procedures for bone densitometry systems employed by the manufacturers, the software algorithms for defining the region of interest or edge detection, and the physiological inhomogeneity of body composition. Whereas intra-unit variation of reproducibility is acceptable, inter-unit variation may reach up to 20%. This paper discusses the problems of designing measurement phantoms and underlines the need for standardisation of phantoms for calibration, cross-calibration, and quality control in bone densitometry. A general phantom used for cross-calibration should handle all parameters influencing measurement of bone minerals to yielded dynamic reference values. One has to note that densitometry systems do not measure the absolute bone mineral content but a model-related equivalent of the calibration material.