Feasibility of low concentration iodinated contrast medium in portal venous abdominal imaging: performance of photon-counting CT virtual monoenergetic imaging compared to conventional EID-CT.
TL;DR
VMI at 45-50 keV provided the optimal balance of contrast and noise, supporting its potential clinical feasibility for routine portal venous phase abdominal imaging, and identifying the optimal keV range.
Abstract
Purpose
To evaluate whether reduced-iodine photon-counting CT (RI-PCCT) virtual monoenergetic imaging (VMI) provides comparable quantitative and non-inferior qualitative image quality compared with standard-iodine energy-integrating detector CT (SI-EIDCT) in portal venous phase abdominal imaging, and to identify the optimal keV range.
Materials And Methods
This retrospective study included 84 adult patients who underwent abdominopelvic PCCT using a lower-concentration iodinated contrast medium (270 mg I/mL) administered at the same injection volume as routine practice, resulting in a reduced total iodine dose, and who had a prior conventional EID-CT performed with standard contrast medium (350 mg I/mL) within one year. Organ-specific contrast-to-noise ratio (CNR) and signal-to-noise ratio (SNR) were calculated across virtual monoenergetic imaging (VMI) levels ranging from 40 to 65 keV and compared with EID-CT. Two independent radiologists assessed subjective image quality at predefined VMI levels (40, 45, and 50 keV) using a 5-point Likert scale. Non-inferiority testing was performed using a predefined margin of -0.5.
Results
The total iodine dose was significantly lower in the PCCT group than in the EID-CT group (400.60 ± 29.82 vs. 517.88 ± 39.46 mgI/kg, P < 0.001). At 40 keV, RI-PCCT demonstrated higher CNR than SI-EIDCT for the pancreas, portal vein, and spleen (all P < 0.001), whereas liver CNR was similar. SNR showed organ-specific reductions, particularly in the liver. Qualitative scores were lower at 40 keV, but 45 keV and 50 keV met the predefined non-inferiority criterion for all qualitative parameters using a -0.5-point margin. Interobserver agreement was excellent for all qualitative parameters (Gwet's AC2, 0.932-0.975).
Conclusions
RI-PCCT provided comparable quantitative and non-inferior qualitative image quality compared with SI-EIDCT, while lowering the iodine and radiation dose. VMI at 45-50 keV provided the optimal balance of contrast and noise, supporting its potential clinical feasibility for routine portal venous phase imaging.