The impact of post-processing images of abdominal CT small vessels using contrast enhancement boost technique: a retrospective study

Abstract

Background: Accurate visualization of small abdominal arteries is crucial for diagnosis, preoperative evaluation, and treatment planning, especially in hepatobiliary and pancreatic diseases. However, due to their small calibers and rapid contrast transit, consistent delineation of these vessels remains challenging in conventional computed tomography (CT) angiography. Recent advancements in image postprocessing, such as contrast enhancement boost (CE-Boost) techniques, offer opportunities to improve vascular conspicuity without additional radiation or contrast agent. This study aimed to evaluate the efficacy of CE-Boost technology in improving image quality and vascular delineation for small abdominal vessels in multiphase contrast-enhanced CT examinations.

Methods: This retrospective analysis included 100 patients undergoing triphasic abdominal CT between July to November 2024. Raw datasets were reconstructed using an adaptive iterative denoising algorithm to generate conventional images (Group A). Subsequent application of flexible subtraction CE-Boost technology produced optimized images (Group B). Quantitative analysis measured CT attenuation values and noise levels (standard deviation) in four arterial branches (common hepatic, left gastric, splenic, and superior mesenteric arteries) and adjacent erector spinae musculature. Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated for vascular structures. Two blinded radiologists independently scored image quality using a 5-point Likert scale, with interobserver agreement assessed via Cohen's kappa.

Results: CE-Boost-processed images (Group B) demonstrated significantly improved image quality compared to conventional processing (Group A). Quantitative analysis showed higher CT attenuation values in Group B across all evaluated arterial branches: common hepatic artery [median 362.50 vs. 256.00 Hounsfield units (HU), P<0.001], left gastric artery (333.87±77.27 vs. 230.38±53.90 HU, P<0.001), splenic artery (median 374.50 vs. 257.17 HU, P<0.001), and superior mesenteric artery (median 380.67 vs. 269.00 HU, P<0.001). Mean increases ranged from 48 to 62 HU. Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) improved by 39-54% across target vessels (all P<0.001). Subjective scoring also favored Group B, with higher overall image quality scores {median: 5 [interquartile range (IQR): 4-5] vs. 4 (IQR: 4-5), P<0.001}, and strong interobserver agreement (κ=0.776 for Group B, κ=0.723 for Group A, both P<0.001). These results indicate superior vascular enhancement and diagnostic visibility using CE-Boost processing.

Conclusions: CE-Boost technology significantly enhances the visualization of small abdominal vasculature through advanced postprocessing optimization. The technique improves objective image quality metrics (CT attenuation, SNR, CNR) while maintaining diagnostic noise levels, and demonstrates high clinical utility for vascular mapping. These advancements in image processing workflows may facilitate more accurate anatomical assessment and diagnostic interpretation of small vessel pathologies.

Keywords: Flexible subtraction; X-ray computed; image quality; tomography.

Figure 1

Comparison of abdominal CT small vessel images before and after CE-Boost post-processing in the same patient. Conventional group (A,C) and CE-Boost (B,D) show the common hepatic artery. (A,C) The artery is blurred with low contrast; (B,D) it’s clearer with enhanced details. Panel (E) (conventional) displays the left gastric artery, while panel (F) (CE-Boost) demonstrates improved vessel clarity, facilitating more detailed morphological analysis. Arrows indicate the common hepatic artery (A-D) and left gastric artery (E,F). CE-Boost, contrast enhancement boost; CT, computed tomography.

Figure 2

Comparison of VR images in the same patient: conventional vs. CE-Boost. Rows show same patient images. (A,C) Conventional VR and (B,D) CE-Boost VR, offering improvements for better anatomical viewing. Arrows highlight key abdominal arterial branches in VR images. They demonstrate CE-Boost’s superiority in 3D anatomical depiction, with enhanced vessel continuity and depth compared to fragmented conventional VR. CE-Boost, contrast enhancement boost; VR, volume-rendered.

Figure 3

Visualization chart of objective evaluation results of image quality. Group A: conventional reconstruction; Group B: CE-Boost processed images. (A) Image noise (standard deviation, HU) measured in adjacent erector spinae musculature. Noise levels were comparable between groups (P=0.152). (B) CT attenuation values (HU) in four arterial branches: CHA, LGA, SA, and SMA. Group B shows significantly higher attenuation (all P<0.001). (C) SNR and (D) CNR for target vessels. Group B demonstrates 39–54% improvement in SNR/CNR (all P<0.001). CE-Boost, contrast enhancement boost; CHA, common hepatic artery; CNR, contrast-to-noise ratio; HU, Hounsfield units; IQR, interquartile range; LGA, left gastric artery; SA, splenic artery; SMA, superior mesenteric artery; SNR, signal-to-noise ratio.