Initial assessment of dual-energy CT in patients with gallstones or bile duct stones: can virtual nonenhanced images replace true nonenhanced images?

Abstract

The purpose of our study was to investigate whether virtual nonenhanced images can replace true nonenhanced images for the evaluation of biliary stones and to compare two virtual nonenhanced images obtained from dual-energy data sets during the late arterial phase (LAP) and portal venous phase (PVP) with true nonenhanced images. One hundred patients with gallstones (n = 77) or bile duct stones (n = 23) who had undergone dual-source dual-energy CT (using 80kVp and 140kVp with the ratio of dose-partitioning, 1:1.3), were included. The CT numbers for the stone, liver, and bile; stone size; and image noise were assessed for each image set. The conspicuity and size of the stones, image quality, and acceptability of the virtual nonenhanced images as a replacement for true nonenhanced images were assessed. Virtual nonenhanced images from both LAP (VNE(LAP)) and PVP showed lower CT numbers for the stone and smaller stone size than did true nonenhanced images (p < 0.05). VNE(LAP) images failed to show stones in 16 patients (16%) with relatively radiolucent bile duct stones (< 78 HU) or small gallstones (< 9 mm(2), < 1.7 mm). There was no difference between the two virtual nonenhanced image sets except for the contrast-to-noise ratio of stone to bile and the image noise (CT numbers for the stone, liver, and muscle; stone size; and contrast-to-noise ratio of stone to liver, p = 1; the CT number for the bile, p = 0.1618; and effective dose, p = 0.7478). VNE(LAP) images were considered acceptable as a replacement for true nonenhanced images in 89% and 90% by reviewers 1 and 2, respectively. Virtual nonenhanced images allow the detection of biliary stones with moderate accuracy, irrespective of the dual-energy contrast-enhanced phase. The detection of gallbladder stones of 9 mm(2) or smaller and of relatively radiolucent bile duct stones of 78 HU or lower is limited.