Evaluation of postprocessing dual-energy methods in quantitative computed tomography. Part 2. Practical aspects.

Three facets of dual-energy quantitative computed tomography are studied: (1) the algorithm for postprocessing data (the methods of Cann, Laval-Jeantet et al, Goodsitt et al [two methods], and Nickoloff et al); (2) the influence of choice of tissue-equivalent materials for calibration; and (3) the difference between central and peripheral calibration. The different tissue-equivalent materials include bone mineral-equivalent (K2HPO4 solutions and calcium hydroxyapatite), fat-equivalent (liquid paraffin, polyethylene, and 70% ethanol solution), and red marrow-equivalent (plastic). Deviation from the manufacturer's quoted content is least with central positioning of the calibration materials. The accuracy of estimates is best when the same tissue-equivalent materials are used for calibration that are being measured. The deviations produced by the use of different tissue-equivalent materials indicate the importance of using materials that mimic the components of bone most closely. The two methods of Goodsitt et al and the method of Nickoloff et al produced the best results.