OBJECTIVE

The purpose of this study is to investigate the magnitude of differences between attenuation values measured on virtual unenhanced images and true unenhanced images obtained using third-generation dual-source dual-energy CT (DECT).

SUBJECTS AND METHODS

A total of 83 patients requiring thoracoabdominal CT for cancer workup were included in this prospective study. CT examinations included true unenhanced acquisitions (tube potential, 120 kVp) and arterial and portal phase dual-energy CT (DECT) acquisitions (tube potential, 100 kVp and Sn 150 kVp [where Sn denotes the interposition of a tin filter in the high-energy beam]; tube current-exposure time product, 190 and 95 mAs). Virtual unenhanced images were created using two commercially available DECT postprocessing algorithms, one of which was designed to create liver images (hereafter referred to as VNC1 images) and the other of which was designed to create images of organs containing minor amounts of fat (hereafter referred to as VNC2 images). Attenuation values on the liver, spleen, paraspinal muscles, retroperitoneal fat, renal cortex and medulla, and gallbladder and bladder lumens were measured.

RESULTS

The attenuation values of all tissues were significantly different between virtual unenhanced and true unenhanced images (p = < 0.001-0.042), except for the liver and spleen in the portal phase and muscles in both phases. When statistically significant, correlations between these differences and body mass index (weight in kilograms divided by the square of height in meters) depended on the tissue imaged and algorithm used. The percentage of cases in which these differences were 10 HU or greater was 1% for the liver and approximately 5% for the spleen and muscles, regardless of the algorithm and phase, but on VNC1 images it reached approximately 30% for the kidney, 70% for the gallbladder and bladder, and depending on the phase, 40-70% for fat. On VNC2 images, the percentage of cases in which these differences were 20 HU or greater was approximately 90% for fat.

CONCLUSION

Abdominal virtual unenhanced images obtained with third-generation dual-source DECT still should not replace true unenhanced images because of substantial differences in attenuation measurements for fluid, fat, and renal tissues.