Low-Energy Spectral CT Sharply Improves Imaging of the Gastrocolic Trunk of Henle

A subtle but surgically crucial vein at the junction of the stomach and colon can now be seen far more clearly on computed tomography, according to a new study testing multi-parameter imaging with dual-layer detector spectral CT. The structure in question, the gastrocolic trunk of Henle, is a short venous trunk that forms where several tributaries converge, and it has long been a hazard zone for surgeons operating in the right upper abdomen. Researchers report that virtual monoenergetic images reconstructed at 40 keV deliver markedly better contrast, signal-to-noise ratio, and subjective image quality than conventional 120 kVp images, potentially giving surgical teams a sharper preoperative map of this variable vessel.

The gastrocolic trunk of Henle typically arises from the confluence of the right gastroepiploic vein, the right colic vein, and the anterior superior pancreaticoduodenal vein, and in some individuals it also incorporates the superior right colic vein or the middle colic vein. Its anatomy varies considerably from person to person, and because the trunk is short, robust, and deeply placed, injury during laparoscopic right hemicolectomy can produce difficult-to-control venous bleeding. The vessel serves as a critical control point in that operation, so accurately characterizing its variations is important for preventing iatrogenic vascular injury and ensuring complete tumor resection. Until now, most knowledge of this anatomy has come from intraoperative observations and cadaveric dissections rather than large-scale in vivo imaging.

Conventional CT can visualize vascular anatomy clearly, accurately, and non-invasively, which makes it attractive for studying the trunk in living patients. Yet the technique has a weakness: as the injected contrast agent dilutes during systemic circulation, the concentration of iodine within the portal venous system falls, and small branches such as the anterior superior pancreaticoduodenal vein become hard to resolve. This dilution limits the anatomical assessment of the gastrocolic trunk via CT venography. Dual-layer detector spectral CT addresses the problem differently. Its stacked detector configuration places an upper layer that absorbs low-energy X-ray photons above a lower layer that captures high-energy photons, allowing simultaneous dual-energy data acquisition and decoupling in a single scan without repeated scanning or voltage switching.

That detector design matters because it enables retrospective generation of multiple image types, including virtual monoenergetic images from 40 to 200 keV, iodine density maps, and effective atomic number maps, without any additional radiation exposure. Virtual monoenergetic imaging has already been applied in arterial imaging, but its use for the gastrocolic trunk had not previously been reported. Low-energy virtual monoenergetic images in the 40 to 60 keV range approximate the K-edge of iodine at 33.2 keV, which amplifies the photoelectric effect and substantially raises the CT values of iodine-containing vessels, helping to reveal distal small vessels that conventional images may miss.

To test the approach, the team retrospectively reviewed 94 patients who underwent abdominal enhanced spectral CT scans between August 2025 and September 2025, ultimately enrolling 61 who met the eligibility criteria. The cohort comprised 35 men and 26 women aged 36 to 81 years, with a mean age of 60.91 years. Patients were included only if the trunk was visible in the portal venous phase, they had no prior gastrointestinal surgery, and their scans were free of respiratory artifacts. Exclusion criteria covered severe image artifacts, a history of abdominal or vascular surgery, tumors obscuring the vessel, and severe hepatic or renal dysfunction. All scanning was performed on a Philips Spectral CT 7500 with a 120 kV tube voltage, 1 mm slices, and iopamidol contrast injected at 3.0 mL per second.

The raw spectral data were reconstructed into conventional 120 kVp images and virtual monoenergetic images at 40, 50, 60, and 70 keV, along with iodine density and effective atomic number maps. Regions of interest, each sized at two-thirds of the vessel lumen diameter to avoid the vascular wall, were placed on the main trunk and on the erector spinae muscle at the same axial level as a background reference. Two senior radiologists measured CT values and standard deviations independently, and mean values were used to calculate signal-to-noise and contrast-to-noise ratios. The same two radiologists also scored volume-rendered images on a five-point scale, where five indicated excellent vessel signal with very sharp edges, and their agreement proved excellent, with Kappa values of at least 0.89.

The quantitative results favored the lowest energy level decisively. For the trunk visualized on 40 keV images, the CT value reached 530.56 plus or minus 82.33 HU, the signal-to-noise ratio 25.34 plus or minus 5.63, and the contrast-to-noise ratio 57.85 plus or minus 12.43, each significantly higher than in every other group with all P values below 0.001. CT values of the vessel rose as energy decreased, and apart from the 70 keV images, all virtual monoenergetic groups showed significantly higher CT values than the conventional images. Image noise moved in the opposite direction: the 40 keV images were the noisiest and conventional images the quietest, with noise falling steadily as energy increased. Even so, the noise on 40 keV images remained less than twice that of conventional images, an outcome the authors attribute to anti-correlation noise suppression technology in the scanner, which curbs noise while preserving the high-contrast advantage of low-energy imaging.

Subjective assessment told the same story. Scores declined as energy levels increased, the 40 keV group earned the highest ratings with clear visualization of the trunk’s anatomical structure and details and good contrast against surrounding tissue, and there was no significant difference in subjective scores between the 70 keV and conventional groups. The iodine density maps showed high-density signals in vascular regions against relatively low-density surroundings, and the effective atomic number maps displayed distinct coloration between vascular and non-vascular areas. These maps cannot undergo three-dimensional reconstruction, so they cannot provide a stereoscopic view of the vessel on their own, but the authors position them as a robust complement to the 40 keV images, with iodine density maps potentially quantifying enhancement and improving the detectability of microvessels.

The study’s authors frame the finding as support for clinical practice and preoperative surgical planning, noting that combining 40 keV virtual monoenergetic images with iodine density and effective atomic number maps allows the branching patterns of the trunk to be visualized more comprehensively. They also situate the result against prior work showing that low-energy virtual monoenergetic images improve visualization of the common hepatic artery, portal vein, and hepatic vein, and that 40 keV imaging has delivered the highest diagnostic performance for portal vein thrombosis and favorable adrenal vein visualization. Not every prior result aligns: studies of lower extremity veins found that 40 keV images were less noisy than conventional images, which the authors attribute to the trunk’s small caliber, making it more susceptible to partial volume effects and uneven contrast filling than large vessels such as the portal vein or aorta.

The researchers are candid about the limits of the evidence. This was a retrospective, single-center study, which makes it vulnerable to selection bias and restricts generalizability, and the anatomical classification of the trunk based on 40 keV images requires validation in a larger sample. They also flag open questions about whether optimizing contrast agent concentration, dose, and injection rate could further improve image quality, and whether low-voltage, low-radiation-dose scanning protocols could substitute for conventional 120 kVp acquisition. Within those boundaries, the study offers a concrete, measured demonstration that a single portal venous phase scan on a dual-layer detector CT, reconstructed at 40 keV and supplemented with material-specific maps, can substantially sharpen the image of a vessel that surgeons would rather map before they encounter it.

Subject of Research: Dual-layer detector spectral CT imaging of the gastrocolic trunk of Henle

Article Title: Multi-parameter imaging of dual-layer detector spectral CT in improving the imaging quality of gastrocolic trunk of henle (GTH)

Article References: He, X., Mu, N., Li, J., Liu, L., Shan, Y., Wang, C., Pan, Y., & Xiao, J. (2026). Multi-parameter imaging of dual-layer detector spectral CT in improving the imaging quality of gastrocolic trunk of henle (GTH). Discover Imaging, 3(1), Article 10. https://doi.org/10.1007/s44352-026-00031-8

Image Credits: AI Generated

DOI: 10.1007/s44352-026-00031-8

Keywords: dual-layer detector CT, spectral CT, virtual monoenergetic images, gastrocolic trunk of Henle, portal venous phase, image quality, iodine density map, effective atomic number, laparoscopic right hemicolectomy, vascular imaging, CT venography, preoperative planning