Preliminary experience with abdominal dual-energy CT (DECT): true versus virtual nonenhanced images of the liver

Purpose. The aim of this work was to compare the quality and noise of true nonenhanced (TNE) and virtual nonenhanced (VNE) images in patients undergoing dual-energy computed tomography (DECT) of the liver. Materials and methods. Twenty consecutive patients (mean age 54.7 +/- 19.9 years) prospectively underwent abdominal DECT to assess the liver using a triphasic protocol consisting of precontrast, arterial-phase and portal-phase acquisitions. Exclusion criteria were allergy to iodinated contrast material, impaired renal function and a body mass index (BMI) >35 kg/m(2). The DE portal-phase acquisition was performed with automatic dose modulation (CARE Dose 4D). Nonionic iodinated contrast material (lomeron 400) was administered at 0.625 gI/kg with a flow rate of 3.5 ml/s. Axial VNE images were reconstructed based on the portal data set using a collimation and an increment of 5 mm and were compared with TNE images reconstructed with the same parameters. The average image quality and noise were analysed by two radiologists in separate reading sessions. Results. No statistically significant difference (p>0.05) in image quality was observed between VNE (4.00 +/- 0.85) and TNE images (4.35 +/- 0.58). A sufficient diagnostic quality was found in 95.0% (19/20) of VNE images and in 100% of TNE images. No statistically significant difference (p<0.05) was observed in the average image noise of VNE (9.5 +/- 0.7) and TNE (12.3 +/- 1.1) images. Conclusions. Abdominal DECT allows acquisition of liver VNE images with similar image quality and lower noise than TNE. Nevertheless, a few technical limitations related to the small field of view of the second detector in patients with a high BMI and heterogeneous iodine subtraction restrict the application of this technique to selected patients only.