Computed tomographic volumetric calculation reproducibility.

Previous studies have described computed tomographic (CT) measurement of the total volume of large organs and tumors. This study examines sources of error in the CT volumetric measurement of small anatomic structures, and has potential application to the volumetric measurement of small extravisceral tumors. Volume is determined by multiplying observer measurement of area by the CT slice thickness. We focused on the measurement of small cross-sectional areas as a major source of error in volumetric calculations. One observer made ten area measurements on each of 12 structures of various sizes. For areas larger than 8 to 10 cm2, any single measurement was within 2 to 4% of the mean of ten values. For areas smaller than 8 to 10 cm2, measurements varied more with respect to the mean, and rose approximately exponentially as the measured cross-sectional area approached zero.