Black-blood computed tomography with contrast-enhancement boost for optimized portal vein imaging and segmentation: a feasibility study

Introduction

Abdominal computed tomography angiography (CTA) remains essential for evaluating vascular pathologies and abdominal diseases (1-4). Given that approximately 80% of hepatic perfusion originates from the portal venous system, accurate and efficient portal vein imaging is crucial for clinical decision-making, particularly in liver transplantation, hepatocellular carcinoma staging, and portal hypertension management (4-7). However, diagnostic quality in portal vein imaging is often diminished by limited contrast between vessels and surrounding tissues (8).

To overcome these limitations, advanced image-processing techniques such as contrast-enhancement boost (CE-boost) have emerged, aiming to enhance vascular conspicuity by combining deformable image registration with subtraction algorithms. CE-boost generates motion-corrected iodine maps by subtracting unenhanced from contrast-enhanced computed tomography (CT) images, compensating for respiratory motion. Subsequently, these iodine maps are superimposed onto original contrast-enhanced CT (CECT) data, amplifying vascular contrast and significantly improving abdominal CTA and portal venous visualization (9,10).

Despite these advancements, manual segmentation of the portal vein remains a tedious, error-prone, and time-consuming task, frequently hindered by suboptimal venous enhancement and residual arterial signals. These limitations particularly affect surgical planning for procedures such as transjugular intrahepatic portosystemic shunt (TIPS), where precise portal vein delineation is vital (7,11). Recent advances in artificial intelligence (AI)-driven image reconstruction and segmentation have shown great promise in improving image quality, noise suppression, and vascular delineation across a wide range of CT applications. Nevertheless, the clinical adoption of AI-assisted segmentation remains limited by the need for transparent models, robust external validation, and consistent reproducibility across imaging workflows (11-13).

Similar to CE-boost, black-blood CT (BBCT) imaging employs subtraction techniques and registration algorithms to effectively remove arterial signals by subtracting arterial phase (AP) images from delayed or non-contrast images (14-16). Although BBCT shows potential in vascular imaging, systematic exploration of its application in portal vein segmentation remains unexplored.

In addition, prior studies have improved portal venous imaging by enhancing portal vein contrast using low-kVp protocols with iterative reconstruction and individualized scan timing with bolus tracking (17,18); however, these techniques do not explicitly suppress residual arterial signal during the portal venous phase (PVP) and therefore may be less effective in isolating the venous system to reduce manual segmentation time.

Accordingly, this study integrated CE-boost technology with BBCT imaging, hypothesizing that this combination will significantly enhance portal vein contrast and streamline segmentation processes. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1813/rc).

Methods

Study population

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Medical Ethics Committee of The First Affiliated Hospital of Sun Yat-sen University (No. [2023]668-1) and the requirement for individual consent for this retrospective analysis was waived.

This retrospective analysis included 50 consecutive patients who underwent abdominal CTA between April and May 2024 at The First Affiliated Hospital of Sun Yat-sen University. The cohort comprised 34 males and 16 females with ages ranging from 21 to 77 years [mean ± standard deviation (SD): 49.68±14.19 years] and an average body mass index (BMI) of 22.54±3.00 kg/m². Although this was a retrospective study, the inclusion and exclusion criteria (e.g., contrast allergy, pregnancy, severe organ dysfunction) were applied to ensure patient safety in the original clinical CT examinations and to maintain consistency with prospective trial designs. These criteria may limit the generalizability but improve reproducibility and reliability of the results. The inclusion criteria were as follows: (I) signed informed consent for CECT examination; and (II) diagnostic-quality PVP images. The exclusion criteria were as follows: (I) history of iodinated contrast allergy; (II) severe cardiac or renal insufficiency; (III) prior hepatic embolization procedures; (IV) pregnancy; and (V) moderate-to-severe hepatic steatosis [liver attenuation <40 Hounsfield units (HU)].

CT scan protocols and image processing

All examinations were performed using a Canon 320-row-detector CT scanner (Aquilion ONE GENESIS, Canon Medical Systems, Tochigi, Japan). A standard abdominal CTA scanning protocol was followed: patients were positioned supine with both arms raised above the head and feet first. The scanning direction was from head to feet, covering from 4–5 cm above the diaphragm to the level of the anterior superior iliac spine. Technical parameters included the following: 120 kV tube voltage with automatic tube current modulation with a target SD of 6.8, 0.275 s/rotation gantry speed, 1 mm slice thickness with 0.8 mm reconstruction interval, and 515×512 matrix. A dual-head high-pressure injector (Medrad Stellant D, Bayer Radiology, Pittsburgh, PA, USA) was used to administer a non-ionic contrast agent, iopromide (370 mgI/mL), through an 18 G intravenous catheter placed in the right antecubital vein. The contrast agent was injected at a rate of 4 mL/s, with a dose of 1.2 mL/kg, followed by a 50 mL saline flush at the same rate. A bolus-tracking technique was implemented with region of interest (ROI) placement in abdominal aorta at hepatic portal level. The ROI for threshold monitoring was set at the level of the abdominal aorta corresponding to the hepatic hilum, with a threshold of 180 HU. Automatic triggering technology was employed; once the threshold was reached, the scanner instructed the patient to hold their breath and initiated scanning to obtain the AP. The PVP was acquired 15 seconds after AP acquisition. The reconstructed images for both the AP and PVP were generated using Adaptive Iterative Dose Reduction Three-Dimensional (AIDR 3D, Canon Medical Systems) with a slice thickness of 1 mm and an interval of 1 mm. The reconstruction kernel used for AIDR-3D was the standard FC08.

The reconstructed PVP images were designated as the control group (Group A). These images were processed using ^SURESubtraction Iodine Mapping software (Canon Medical Systems) to generate CE-boost PVP images (Group B). In this step, the non-contrast dataset was registered to the PVP dataset, subtracted to obtain an iodine map representing the enhanced signal, and then superimposed onto the PVP image to improve portal venous contrast. Subsequently, to suppress residual arterial enhancement amplified by CE-boost, Group B images were registered with the AP images and subtracted to produce the final CE-boost + BBCT images (Group C). A physician with 12 years of experience in image post-processing independently extracted the portal venous system from the three sets of images and rendered the vascular structures using multiplanar maximum intensity projection (MIP) and volume rendering (VR). To minimize potential learning effects, the three groups (A, B, and C) were processed with a 2-week interval between sessions. All portal venous segmentations were performed on Vitrea workstation (version 6.9.2, Canon Medical Systems). The radiation dose parameters and post-processing time required for each group were also recorded.

Subjective image quality analysis

Due to the distinctive features exhibited by the black-blood processed multiplanar reconstruction (MPR) images, a blinded evaluation was performed to obscure patients’ clinical information and post-processing techniques. Accordingly, two senior radiologists with 8 and 10 years of experience independently performed subjective scoring of the portal venous system as extracted from MIP and VR images. This post-processing provided a three-dimensional display of the portal venous system, including the main trunk and branches of the portal vein, as well as the superior mesenteric and splenic veins. The image quality scoring system assessed four key parameters: (I) number of visible portal venous branches; (II) vascular wall delineation; (III) artifact severity; and (IV) noise levels, graded on a 5-point scale with scores of 3 and above deemed acceptable for diagnostic purposes, and scores below 3 considered non-diagnostic. The scoring system was as follows:

• 1 point: extremely poor image quality; main portal vein poorly visualized; left and right branches and distal small branches not visible; excessive artifacts; very high noise.
• 2 points: poor image quality; main portal vein moderately visualized; left and right branches poorly visualized; distal small branches not visible; numerous artifacts; high noise.
• 3 points: average image quality; main portal vein clearly visualized; left and right branches moderately visualized; distal small branches poorly visualized; moderate artifacts and noise.
• 4 points: good image quality; main portal vein and left and right branches clearly visualized; distal small branches moderately clear; minimal artifacts and noise.
• 5 points: excellent image quality; main portal vein and left and right branches clearly visualized; distal small branches clearly visualized; no artifacts; very low noise.

Objective image quality analysis

All PVP images were quantitatively analyzed using axial source images with standardized window settings (width 500 HU, level 70 HU). ROIs were systematically placed at predefined anatomical landmarks including ambient air, psoas major muscle, abdominal aorta, hepatic artery, segment VII liver parenchyma, inferior vena cava, main portal vein, and bilateral portal vein branches. A circular ROI template of 100 mm² was applied, with adaptive resizing for vessels smaller than 100 mm² to maximize lumen coverage while maintaining a 1 mm buffer zone from vessel walls and excluding calcifications. Attenuation values (HU) and SD were recorded for each ROI. Signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated using the following formulas:

SNR=CTvessel/SDvesselCNR=(CTvessel−CTmuscle)/SDair[1]

Statistical analysis

Statistical analyses were performed using the software SPSS 26.0 (IBM Corp., Armonk, NY, USA). Quantitative data were expressed as mean ± SD, and categorical variables as counts and percentages. Normality of continuous variables was assessed using the Shapiro-Wilk test. Quantitative parameters (CT value, SNR, and CNR) that followed a normal distribution were compared among the three groups using one-way analysis of variance (ANOVA) with Bonferroni-adjusted post hoc comparisons. Subjective image quality scores, being ordinal variables, were expressed as median [interquartile range (IQR)] and analyzed using the Kruskal-Wallis test followed by Dunn-Bonferroni correction for pairwise comparisons. Inter-observer agreement between the two radiologists was evaluated using the intraclass correlation coefficient (ICC) with a two-way random-effects model. An ICC value of 0.81–1.00 indicated excellent agreement, 0.61–0.80 good agreement, 0.41–0.60 moderate agreement, and ≤0.40 poor agreement. For all analyses, a two-sided P value <0.05 was considered statistically significant.

Results

Radiation dose

The mean dose-length product (DLP) for arterial and PVPs was 155.59±43.61 mGy·cm, and the mean effective dose was 7.00±1.96 mSv (k=0.015).

Post‑processing time

The black-blood CE-boost (Group C) [2.2 (1.9, 3.5) min] required significantly less post-processing time than both control (Group A) [5.1 (3.6, 7.9) min] and CE-boost groups (Group B) [4.2 (2.8, 6.7) min] (P<0.001), as it inherently eliminates the need for separation and processing of residual arterial artifacts during the PVP.

Subjective image evaluation

For all three groups, the interobserver agreement was excellent, with ICC values of 0.926 [95% confidence interval (CI): 0.869–0.958, P<0.001], 0.938 (95% CI: 0.890–0.965, P<0.001), and 0.922 (95% CI: 0.862–0.956, P<0.001), respectively. Group C achieved significantly higher median subjective scores for portal vein image quality compared to both conventional and standard CE-boost sequences [5.0 (IQR, 4.9, 5.0) vs. 3.7 (3.7, 3.7) vs. 4.4 (4.3, 4.4); P<0.001], as detailed in Table 1.

Objective image evaluation

The quantitative analysis of CT values, SD, SNR, and CNR across the groups is summarized in Table 2. Statistically significant differences (P<0.001) were observed among the three groups in CT values, SNR, and CNR for the liver parenchyma, inferior vena cava, main portal vein, and bilateral hepatic portal veins. Group C demonstrated superior SNR and CNR compared to both Groups A and B (P<0.001). Notably, group C exhibited slightly higher SD values in the liver parenchyma and left portal vein due to cumulative noise from image superposition (liver parenchyma: P=0.04; left portal vein: P=0.074). For all other anatomical regions, intergroup differences in CT values, SD, SNR, and CNR reached statistical significance (P<0.05).

Discussion

This study demonstrates the effectiveness of integrating CE-boost and BBCT techniques to significantly enhance portal venous image quality and streamline post-processing workflows. To the best of our knowledge, this is the first report evaluating black-blood imaging’s impact on portal vein image quality and segmentation efficiency.

Our results indicated that combined CE-boost and BBCT processing (Group C) substantially improved both subjective image quality scores and objective metrics (SNR and CNR) compared with Groups A and B (Table 2). These findings align with prior studies, underscoring CE-boost’s capability to enhance vascular contrast in abdominal CTA (9,10). Additionally, incorporating BBCT notably elevated overall image clarity by selectively suppressing arterial signals that often complicate portal vein delineation during manual segmentation (14,15,19). Our approach uniquely targeted arterial signal suppression, facilitating isolated portal vein assessment.

3D vascular reconstructions are integral for accurate diagnosis, surgical planning, and therapeutic interventions (20-22). However, the complex hepatic vasculature often causes misclassification during manual segmentation (7,8,12). By effectively eliminating residual arterial enhancement, BBCT streamlined the segmentation workflow, leading to shorter processing times and improved consistency (Figures 1,2). Although the reduction was modest on a per-case basis, such time savings may accumulate considerably in high-volume centers, thereby enhancing overall workflow efficiency.

Figure 1 A 57-year-old woman with two weeks of vague upper abdominal discomfort underwent portal venous CTA. (A1,B1,C1) The PVP. (A2,B2,C2) The initial workstation interface. (A3,B3,C3) 3D volume-rendered images of the portal vein. (A4,B4,C4) MIP images of the portal vein. (A1-A4) The original image. (B1-B4) CE-boost processed image. (C1-C4) CE-boost with black-blood technique. Manual post-processing times were 4 min 3 sec for the original (A3), 4 min 24 sec for CE-boost (B3), and 2 min 32 sec for CE-boost with black-blood subtraction (C3). (B1-B4) CE-boost improved portal vein attenuation and revealed more branches but retained arterial signals, complicating isolated portal venous visualization. (C1-C4) The black-blood subtraction technique effectively suppressed arterial signals, reduced segmentation time, and clarified portal vein branches, enhancing diagnostic assessment. 3D, three-dimensional; CE-boost, contrast-enhancement boost; CTA, computed tomography angiography; MIP, maximum intensity projection; PVP, portal venous phase.

Figure 2 A 55-year-old male with cholangiocarcinoma underwent portal venous CTA. (A) The PVP. (B) CE-boost processed image. (C) CE-boost with black-blood technique. The main portal vein’s mean CT attenuation increased from 93.9 HU (A) to 126 HU (B) and 234 HU (C). CE-boost enhances vascular contrast, and black-blood subtraction further brightens the portal vein, improving tumor invasion assessment. CE-boost, contrast-enhancement boost; CT, computed tomography; CTA, computed tomography angiography; HU, Hounsfield units; PVP, portal venous phase; Std, standard.

Despite these promising outcomes, the study has limitations. Firstly, its retrospective, single-center design inherently restricts the generalizability of the findings. Secondly, the relatively small sample size limits statistical robustness. Thirdly, the CE-boost and BBCT techniques are platform-dependent implementations based on Canon’s subtraction algorithm, which may limit their applicability on other CT systems. Nevertheless, the processed datasets can be exported to different post-processing workstations (e.g., SyngoVia, Siemens, Erlangen, Germany; CareStream, Carestream Health, Rochester, NY, USA) for portal vein segmentation and analysis. Variations among these platforms may influence segmentation time and workflow efficiency, which should be compared in future studies. In addition, segmentation was performed by a single experienced radiologist to ensure consistency. Multi-reader validation will be explored in future studies. Lastly, the absence of direct evaluation of clinical outcomes necessitates further investigation. Future prospective studies, preferably multicentered with larger populations, should validate these preliminary findings. Moreover, although improved portal vein delineation is promising, its direct impact on surgical outcomes requires further investigation. Comparisons with state-of-the-art AI-based segmentation methods should also be considered in future studies.

Conclusions

Combining CE-boost with BBCT is a feasible technique that improves portal vein image quality and reduces manual segmentation time.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1813/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1813/dss

Funding: This study was funded by the National Natural Science Foundation of China (No. 82202217).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1813/coif). R.X. is an employee of Canon Medical Systems. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Medical Ethics Committee of The First Affiliated Hospital of Sun Yat-sen University (No. [2023]668-1) and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Chernyak V, Fowler KJ, Kamaya A, Kielar AZ, Elsayes KM, Bashir MR, Kono Y, Do RK, Mitchell DG, Singal AG, Tang A, Sirlin CB. Liver Imaging Reporting and Data System (LI-RADS) Version 2018: Imaging of Hepatocellular Carcinoma in At-Risk Patients. Radiology 2018;289:816-30. [Crossref] [PubMed]
2. He W, Xu P, Zhang M, Xu R, Shen X, Mao R, Li XH, Sun CH, Zhang RN, Lin S. Deep learning-based reconstruction improves the image quality of low-dose CT enterography in patients with inflammatory bowel disease. Abdom Radiol (NY) 2025;50:3907-16. [Crossref] [PubMed]
3. Cai H, Jiang H, Xie D, Lai Z, Wu J, Chen M, Yang Z, Xu R, Zeng S, Ma H. Enhancing image quality in computed tomography angiography follow-ups after endovascular aneurysm repair: a comparative study of reconstruction techniques. BMC Med Imaging 2024;24:162. [Crossref] [PubMed]
4. Hagen F, Peisen F, Spogis J, Mair A, Nikolaou K, Stenzl A, Kruck S, Bedke J, Kaufmann S, Thaiss WM. Is a single portal venous phase in contrast-enhanced CT sufficient to detect metastases or recurrence in clear cell renal cell carcinoma? - a single-center retrospective study. Cancer Imaging 2022;22:9. [Crossref] [PubMed]
5. Liu Z, Shen T, Xia K, He J, Rui T, Chen W. Classification of variant portal vein anatomy based on three-dimensional CT: surgical implications. Surg Radiol Anat 2024;46:1177-84. [Crossref] [PubMed]
6. Sureka B, Patidar Y, Bansal K, Rajesh S, Agrawal N, Arora A. Portal vein variations in 1000 patients: surgical and radiological importance. Br J Radiol 2015;88:20150326. [Crossref] [PubMed]
7. Zhao N, Shi A, Huang W, He J, Wang D, Zhang Y, Guo H, Zhou B, Gan H, Pang P. Three-dimensional volumetric CT image fusion and trans-abdominal US: Adjunct guidance to portal vein cannulation for TIPS. Eur J Radiol 2025;183:111875. [Crossref] [PubMed]
8. Matsuda I, Hanaoka S, Akahane M, Sato J, Komatsu S, Inoh S, Kiryu S, Yoshioka N, Ino K, Ohtomo K. Adaptive statistical iterative reconstruction for volume-rendered computed tomography portovenography: improvement of image quality. Jpn J Radiol 2010;28:700-6. [Crossref] [PubMed]
9. Hou J, Zhang Y, Yan J, Zhang T, Xia W, Zhu Y, Liu W. Clinical application of the contrast-enhancement boost technique in computed tomography angiography of the portal vein. Abdom Radiol (NY) 2023;48:806-15. [Crossref] [PubMed]
10. Xu J, Wang S, Wang X, Wang Y, Xue H, Yan J, Xu M, Jin Z. Effects of contrast enhancement boost postprocessing technique in combination with different reconstruction algorithms on the image quality of abdominal CT angiography. Eur J Radiol 2022;154:110388. [Crossref] [PubMed]
11. Li S, Li XG, Zhou F, Zhang Y, Bie Z, Cheng L, Peng J, Li B. Automated segmentation of liver and hepatic vessels on portal venous phase computed tomography images using a deep learning algorithm. J Appl Clin Med Phys 2024;25:e14397. [Crossref] [PubMed]
12. Chierici A, Lareyre F, Salucki B, Iannelli A, Delingette H, Raffort J. Vascular liver segmentation: a narrative review on methods and new insights brought by artificial intelligence. J Int Med Res 2024;52:3000605241263170. [Crossref] [PubMed]
13. Kufel J, Bargieł-Łączek K, Kocot S, Koźlik M, Bartnikowska W, Janik M, Czogalik Ł, Dudek P, Magiera M, Lis A, Paszkiewicz I, Nawrat Z, Cebula M, Gruszczyńska K. What Is Machine Learning, Artificial Neural Networks and Deep Learning?-Examples of Practical Applications in Medicine. Diagnostics (Basel) 2023;13:2582. [Crossref] [PubMed]
14. Lin Y, Liu T, Hu Y, Xu Y, Wang J, Guo S, Xie S, Sun H. Assessment of vascular invasion of pancreatic ductal adenocarcinoma based on CE-boost black blood CT technique. Insights Imaging 2024;15:293. [Crossref] [PubMed]
15. Li J, Zhang Y, Hou J, Li Y, Zhao Z, Xu M, Liu W. Clinical Application of Dark-blood Imaging in Head and Neck CT Angiography: Effect on Image Quality and Plaque Visibility. Acad Radiol 2024;31:2478-87. [Crossref] [PubMed]
16. Zhou T, Li Y, Sun N, Hua B, Zhang M, Ma Z, Li M, Jia S. Superiority of black-blood computed tomography over conventional computed tomography angiography in the identification of soft plaques: a prospective study. Quant Imaging Med Surg 2025;15:6287-300. [Crossref] [PubMed]
17. Ren Z, Zhang X, Hu Z, Li D, Liu Z, Wei D, Jia Y, Yu N, Yu Y, Lei Y, Chen X, Guo C, Ren Z, He T. Reducing Radiation Dose and Improving Image Quality in CT Portal Venography Using 80 kV and Adaptive Statistical Iterative Reconstruction-V in Slender Patients. Acad Radiol 2020;27:233-43. [Crossref] [PubMed]
18. Valletta R, Bonatti M, Vingiani V, Corato V, Proner B, Lombardo F, Avesani G, Pertner P, Zamboni GA. Feasibility of a single-phase portal venous CT protocol using bolus tracking technique and lean body weight-based contrast media dose. Eur Radiol 2025;35:1067-75. [Crossref] [PubMed]
19. Lu Y, Cao R, Jiao S, Li L, Liu C, Hu H, Ma Z, Jiang Y, Chen J. A novel method of carotid artery wall imaging: black-blood CT. Eur Radiol 2024;34:2407-15. [Crossref] [PubMed]
20. Nayantara PV, Kamath S, Manjunath KN, Rajagopal KV. Computer-aided diagnosis of liver lesions using CT images: A systematic review. Comput Biol Med 2020;127:104035. [Crossref] [PubMed]
21. Sakamoto T. Roles of universal three-dimensional image analysis devices that assist surgical operations. J Hepatobiliary Pancreat Sci 2014;21:230-4. [Crossref] [PubMed]
22. Agarwal A, Jain M. Multidetector CT portal venography in evaluation of portosystemic collateral vessels. J Med Imaging Radiat Oncol 2008;52:4-9. [Crossref] [PubMed]