Purpose: To measure scanner and patient variation in computed tomographic (CT) numbers for electron-beam CT and to determine the ability of calibration phantoms to reduce variability in calcium quantitation.
Materials and methods: Two calibration phantoms were imaged to ensure longitudinal homogeneity and to determine the short-term intrascanner variation in CT numbers. Each phantom set was imaged twice a day for 14 weeks to determine intra- and interscanner variation. Data from examinations of 167 patients that included the phantom were analyzed to determine the intra- and interpatient variation in CT numbers of objects with known calcium concentrations.
Results: The calibration reduced scanner variations by approximately 25%. The calcium concentration associated with a CT number of 130 HU varied from 77.1 to 136.4 mg/cm3 and was dependent on patient girth, sex, smoking history, and image level.
Conclusion: Scanner and patient variations in CT numbers in electron-beam CT can be reduced with a calibration phantom. In vitro and in vivo estimates of calcium concentration had a precision of 2% and 7%, respectively.