The study employed a phantom-based experimental approach to evaluate the attenuation of Gd-EOB-DTPA in PCD-CT at varying concentrations and energy levels. Two phantoms were used: a custom-built phantom containing two sets of 20 centrifuge tubes filled with Gd-EOB-DTPA solutions ranging from 0.00 to 2.5 µmol/ml in 0.125 µmol/ml increments, and an anthropomorphic abdominal phantom with a spectral CT core (see Figure 1). The latter was scanned with solutions prepared in 0.25 µmol/ml increments to better simulate clinical conditions. All samples were prepared using sterile pipetting techniques and verified through inductively coupled plasma-optical emission spectroscopy to ensure precise concentrations.
The custom-built phantom scans were performed with intentionally high radiation doses to minimize image noise, using a clinically approved PCD-CT system (NAEOTOM Alpha, Siemens Healthineers). The anthropomorphic phantom was scanned with a standard abdominal protocol, focusing on clinically relevant dose levels. Image reconstructions were performed at 40 keV, 50 keV, 60 keV and 70 keV, along with virtual non-contrast (VNC) series.
Image analysis was conducted on dedicated post-processing software (syngo.via), where Hounsfield unit measurements were obtained by placing circular ROIs in the center of each tube across all reconstructions. To ensure comparability, ROI size and placement remained consistent across measurements. The stepwise increase in attenuation with rising Gd-EOB-DTPA concentrations was quantified across different energy levels.
To estimate hepatic enhancement, the measured attenuation values were extrapolated using previously published pharmacokinetic data. These calculations included adjustments for hepatic uptake rates, particularly at doses below 100 µmol/kg, where biliary excretion of Gd-EOB-DTPA is known to be more efficient and at doses of 200-500 µmol/kg were partial saturation of hepatic transporters was observed5–7. Linear regression models were applied to the phantom measurements to estimate liver attenuation for various doses. The hepatic contrast enhancement CE for the different concentrations derived by the regression were than determined as the delta between 40 keV and VNC images for each concentration subtracted from the delta of the negative sample (pure water) as follows:
CE (y)=40keV(y)-VNC(y))-(40keV(Water)-VNC(Water)
with y as concentration of Gd-EOB-DTPA/ml.
The contrast-to-noise ratio (CNR) was calculated for each concentration to account for image noise, in the anthropomorphic phantom, where noise was significantly higher compared to the custom-built phantom. All scans were repeated to ensure reproducibility, and paired t-tests were conducted to compare the results of both phantoms. The study design allowed for robust assessment of Gd-EOB-DTPA performance, highlighting its potential in dedicated liver imaging with PCD-CT.