Inaccuracies in quantitative CT (QCT) for vertebral bone mineral measurements may result from differences between the temperature of the vertebrae and the calibration standards. This study aims to quantify these effects by using scans of marrow-equivalent materials and computer simulation.Materials and Methods
The CT numbers of fat, water, gelatin suspension, and bone were measured within an anthropomorphic phantom at temperatures between −18 and 38αC. A computer simulation was then performed using these materials to represent marrow fat, soft tissue, and bone in varying proportions over this temperature range. Postprocessing single and dual energy QCT calculations were then performed on the data acquired from the simulation.Results
A change of 80 HU in the CT number of water on cooling from 38 to −18°C was demonstrated. An increase of 95 HU in the CT number of fat occurred over the same temperature range. Dry cortical bone showed no change in CT number with temperature changes from 24 to - 18°C. In the computer simulation, the fat error associated with single energy QCT for trabecular bone mineral densitometry was 20% less for specimens at room temperature than at body temperature. In simulated frozen specimens, varying marrow fat/soft tissue composition had almost no effect on single energy QCT mineral densitometry. Dual energy QCT methods that use a fat-equivalent reference material were significantly influenced by the temperature of the specimen.Conclusion
The fat error of single energy QCT for mineral densitometry may have been underestimated in previous in vitro studies using vertebral specimens scanned at room temperature. In the simulation, the fat error diminished as the temperature of the specimen was reduced and was negligible when frozen. Fat-equivalent reference materials used for dual energy QCT in vivo should have similar X-ray-attenuating properties at room temperature to those of marrow fat at body temperature.