The effect of lipid content on virtual non-contrast images and iodine density estimation from dual-layer spectral...

Dual-layer spectral computed tomography (DLCT) enables virtual non-contrast (VNC) imaging and iodine quantification, and is increasingly used in abdominal imaging. However, lipid deposition in the abdominal organs may interfere with spectral decomposition accuracy, potentially affecting both VNC attenuation and iodine density measurements. This study aimed to evaluate the effect of lipid content on VNC attenuation and iodine density in DLCT.

The concentration of the lipid solution was varied from 0% to 40% using water and water-soluble cutting oils. Five tubes were contrast vials containing a contrast agent, and another five tubes were non-contrast vials without a contrast agent. A water-filled phantom was placed around the tubes. Images were acquired with a DLCT scanner using tube voltages of 120 and 140 kVp. True non-contrast (TNC), VNC, and iodine density images were reconstructed. The VNC estimation error was calculated as the difference in attenuation between the VNC and TNC images. The accuracy of the iodine density was also evaluated.

Increasing the lipid content resulted in a larger VNC estimation error and a greater underestimation of the iodine density at both tube voltages. The measurements obtained at 140 kVp demonstrated smaller VNC estimation errors and less underestimation of the iodine density than those obtained at 120 kVp.

Increased lipid content reduced the accuracy of VNC attenuation and iodine density measurements in DLCT. The use of a higher tube voltage may improve the measurement accuracy compared with lower tube voltage settings.

High-voltage scanning improves the accuracy of VNC attenuation and iodine density analysis interfered by lipid content. These findings can contribute to abdominal imaging, including adrenal adenoma characterization, fatty liver assessment, and extracellular volume evaluation.