Purpose: To assess CT numbers and noise along the z-axis of a wide volumetric Dual Energy CT (DECT) acquisition.
Methods: An 18-cm long phantom was scanned with two consecutive rotations (135/80 kVp) using a 16-cm-wide beam. Material-equivalent rods were inserted in the phantom: adipose, brain white matter, chronic blood, acute blood, and five concentrations of iodine (0.2, 0.5, 1.0, 2.0, 5.0 mg/mL). The following DECT image series were reconstructed: 120 kV equivalent (kVe), 66 keV virtual monoenergetic (VM), iodine map (IM), and virtual non-contrast (VNC) images. CT number and noise were measured in Regions of Interest for each material in 3-mm thick slices along the full scan range (z=0-160 mm). A tolerable scan range was defined for each material and image series to include only “accurate” CT numbers defined within ±5% or ±5 HU of the central CT number (z=80 mm), whichever was greatest.
Results: At the periphery of the phantom, CT numbers measured in the negative thousands for all materials due to cone beam effects. For all materials, average tolerable ranges for the kVe, VM, IM, and VNC image series were 118.7 mm, 120.3 mm, 109.0 mm, and 115.3 mm, respectively. The tolerable range was smallest (105.0 mm) for chronic blood and 0.5 mg/mL iodine in IM and VNC series and largest (132.0 mm) for adipose in VM series. The greatest increase in noise was for iodine 5 mg/mL with 8.88 HU (from 7.4 to 16.28 HU in IM series) along the entire scan range and 3.70 HU (from 11.0 to 14.7 HU in VNC series) within the tolerable range.
Conclusion: For all images, none of the materials maintained accurate CT numbers beyond 132 mm, and some were only accurate within 105 mm. Diagnostic accuracy and material characterization may be affected at the peripheral images in volumetric DECT images.