Masatoshi Hori MD, Presenter: Nothing to Disclose

Kenji Suzuki PhD, Abstract Co-Author: Nothing to Disclose

Mark L. Epstein MS, Abstract Co-Author: Nothing to Disclose

Richard L. Baron MD, Abstract Co-Author: Nothing to Disclose

Although the use of thinner CT slices may improve the accuracy of CT liver volumetry because of reduced partial volume effects, the increased number of images produces a larger workload. No studies have shown systematically a quantitative estimate for the effects of slice thickness on the accuracy of volume calculations and it is unclear how thin CT images should be for optimal use in clinical routines. In this study, we evaluated the relationship between slice thickness and calculated liver volume on CT liver volumetry for living-related liver transplantation.

Twenty adult potential liver donors (12 men, 8 women; mean age, 39 years; range, 24-64 years) underwent CT with a 64-channel multidetector-row CT scanner after intravenous injection of 150 mL of contrast material (300 mgI/mL). Four image sets with slice thicknesses of 0.625 mm (isotropic), 2.5 mm, 5 mm, and 10 mm were used. First, we applied our semi-automated liver extraction software to the four sets of portal venous-phase CT images with different slice thicknesses to obtain the initial liver boundaries. An abdominal radiologist reviewed and edited the initial boundaries on all images to optimize accuracy. Finally, liver volumes were determined by counting of the voxels within the liver boundary. Calculated volumes were compared among the four slice thicknesses.

The mean liver volumes estimated with CT liver volumetry were 1322.5 ± 259.5 cm3 on 0.625-mm, 1313.3 ± 257.8 cm3 on 2.5-mm, 1310.3 ± 260.0 cm3 on 5-mm, and 1268.2 ± 256.8 cm3 on 10-mm images. The volumes calculated for 0.625-mm images were significantly larger than those for thicker images (Dunnett pairwise multiple comparisons t test, P<.0001 for all comparison pairs).

Liver volumes calculated on 2.5-mm or thicker images were significantly smaller than volumes calculated on 0.625-mm-thick 3D isotropic images. However, if a maximum error of 2% in the calculated liver volume is permitted, 5-mm-thick images are acceptable for CT liver volumetry, and 3D isotropic images are not required.

The accuracy of CT liver volumetry can be improved by using thinner slices. However, if a maximum error of 2% is permitted, 5-mm-thick images are acceptable, and 3D isotropic images are not required.

Hori, M, Suzuki, K, Epstein, M, Baron, R, CT Liver Volumetry: Effects of Slice Thickness on Volume Calculation— Can 3D Isotropic CT Data Improve the Accuracy?.  Radiological Society of North America 2010 Scientific Assembly and Annual Meeting, November 28 - December 3, 2010 ,Chicago IL. http://archive.rsna.org/2010/9001852.html