This retrospective study of 40 patients evaluated whether virtual monoenergetic images (VMI) reconstructed from portal venous PCCT could provide sufficient arterial vessel contrast and potentially serve as an alternative when a dedicated arterial phase is unavailable or suboptimal.
40-keV VMI produced the highest arterial attenuation, reaching 439 ± 97 HU versus 360 ± 97 HU on true polychromatic arterial images, corresponding to approximately 18% higher vessel attenuation. This gain came at the cost of increased noise and substantially lower CNR and SNR.
Subjectively, 40 keV provided the best vascular contrast, whereas 70 keV provided the best overall image quality and represented a more balanced reconstruction with improved noise, CNR, and SNR compared with 40 keV. The findings suggest low-keV VMI may help salvage arterial information from portal venous PCCT datasets and potentially avoid repeat examinations in selected cases.
Access publication
Figure 2. The same CT slice with pv contrast (center 130, width 550) at 40 keV (a), 70 keV (b), 100 keV (c), and 190 keV (d). (a) demonstrates major image noise with acceptable image quality and optimal vessel contrast. (b) shows no noise with a good vessel assessability and optimal image quality. (c) shows no noise, acceptable quality, and poor vascular contrast. (d) demonstrates no noise, no vessel contrast, and poor image quality.
Protocol
Detail | Value |
|---|
Scanner | NAEOTOM Alpha |
Application | Contrast-enhanced abdominal PCCT / abdominal arterial vessel assessment |
Study design | Retrospective |
Study population | 40 patients |
Acquisition phases | Arterial and portal venous |
Scan direction | Cranio-caudal |
Tube voltage | 120 kV |
Pitch | 0.80 |
Reconstruction increment | 0.70 mm |
Slice thickness | 1.0 mm |
Reconstruction kernel | Qr40 |
Postprocessing | Monoenergetic+ |
Workstation | Syngo.via VB60A_HF04 |
Contrast agent | Xenetix 350 (iobitridol) |
Contrast volume | 100 mL |
Saline chaser | 50 mL |
Injector | CT Motion XD8000 |
Arterial bolus triggering | CARE Bolus, threshold 140 HU |
Arterial phase delay | 7 s after threshold |
Portal venous phase delay | 55 s |
Analyzed vessels | Descending aorta, coeliac trunk, superior mesenteric artery, bilateral renal arteries, bilateral common iliac arteries |
Objective endpoints | Vessel attenuation, noise, SNR, CNR |
Subjective endpoints | Overall image quality, arterial vessel contrast, image noise using 5-point Likert scales |
Key Results
Parameter | Main finding |
|---|
Arterial attenuation | 40-keV portal venous VMI: 439 ± 97 HU vs 360 ± 97 HU for true arterial polychromatic images (p < 0.001) |
Contrast gain at 40 keV | Approximately 18% higher arterial attenuation than the true arterial phase |
Best vascular contrast | 40 keV provided the highest objective attenuation and best subjective arterial vessel contrast |
Noise | Highest at 40 keV and progressively decreased with increasing energy; VMI at ≥70 keV generally showed lower noise than arterial polychromatic images |
CNR | True arterial images: 25.10 ± 9.38; highest portal venous VMI CNR at 60 keV: 8.64 ± 3.58 |
SNR | True arterial images: 20.10 ± 8.30; all portal venous VMI levels had significantly lower SNR |
Best overall image quality | 70 keV |
Reader agreement | Substantial agreement for vessel contrast (κ = 0.799), image quality (κ = 0.742), and noise (κ = 0.712) |
Main implication | 40-keV VMI can substantially enhance arterial contrast from portal venous PCCT, potentially allowing assessment of arterial vessels after suboptimal contrast timing or reducing the need for repeat/dedicated arterial acquisitions in selected cases. |
Source:
Dillinger D, Overhoff D, Ayx I, Kaatsch HL, Hagen A, Schönberg SO, Waldeck S. Optimizing Arterial Vessel Contrast in Portal Venous Phase with Virtual Monoenergetic Images from Photon-Counting Detector CT Scans of the Abdomen—First Clinical Experiences. Diagnostics. 2024;14(6):627. doi: 10.3390/diagnostics14060627.
