Comparison of Virtual Unenhanced Images Derived From Dual-Energy CT With True Unenhanced Images in Evaluation of Gallstone Disease.

Abstract

The aim of this study was to compare gallstones on virtual unenhanced images and true unenhanced images acquired with dual-energy CT (DECT). We enrolled 112 patients with right upper quadrant pain and clinically suspected acute cholecystitis or gallstone who underwent DECT--including unenhanced, arterial, and portal phases. Eighty-three gallstones with composition proven by semiquantitative Fourier transform infrared spectroscopy from 45 patients who had undergone cholecystectomy (40 cholesterol gallstones from 21 patients, 43 calcium gallstones from 24 patients) were included. CT images were retrospectively evaluated for stone size, contrast-to-noise ratio (CNR) of gallstone to bile, and visibility and density of gallstones for each image set. The visibility of each type of stone was compared with a paired t test. Both cholesterol and calcium stones measured smaller on virtual unenhanced images than on true unenhanced images, yielding a lower sensitivity of virtual unenhanced images for detecting small gallstones. Mean CNR of cholesterol stones was 2.45 ± 1.32 versus 1.67 ± 1.55 (p < 0.032) and that of calcium stones was 10.59 ± 7.15 and 14.11 ± 9.81 (p < 0.001) for virtual unenhanced and true unenhanced images, respectively. For calcium stones, two readers found 43 of 43 (100%) on true unenhanced images; one reader found 41 of 43 (95%) and the other, 37 of 43 (86%) on virtual unenhanced images. For cholesterol stones, one reader found 20 of 40 (50%) and the other 19 of 40 (47%) on true unenhanced images versus 34 of 40 (85%) and 30 of 40 (75%), respectively, on virtual unenhanced images. The visibility of cholesterol stones was higher on virtual unenhanced images, but that of calcium stones was lower. Virtual unenhanced images at DECT allow better visualization of cholesterol gallstones, but true unenhanced images allow better visualization of calcium and small gallstones.