COMPUTED-TOMOGRAPHY OF THE LARGE-BOWEL WALL - CHOICE OF SLICE THICKNESS AND INTRALUMINAL CONTRAST-MEDIUM

RATIONALE AND OBJECTIVES. TO determine the relative contribution of sl ice thickness and intraluminal contrast density to imaging of the larg e bowel wall. METHODS. The authors used phantom experiments to evaluat e the partial volume averaging of 2-, 4-, and 8-mm slices on imaging a single density interface (320/-100 HU) and 5 dual-density interfaces (320/50/-100 HU, 160/50/-100 HU, 0/50/-100 HU, -100/50/100 HU, and -10 00/50/100 HU). For the experiments with a dual-density interface, whic h simulated the bowel wall, the phantom was scanned at an angle 45 deg rees to the scan plane. RESULTS. The most accurate display of interfac es was obtained with a slice thickness of 2 mm. When using 8-mm slices , walls can appear to be much thinner or thicker than normal or not ev en recognizable as a wall depending on the densities in the interface. Adjusting the density of the intraluminal contrast medium to within t he range of the mean and window setting will improve the representatio n of the bowel wall. CONCLUSIONS. Slice thickness has a significant in fluence on the true representation of the bowel wall using computed to mography. This effect is greater than the effect of the intraluminal c ontrast medium density.