TOPLINE

Photon-counting detector CT (PCD-CT) with a sharp kernel and large matrix size significantly improved the detection of visceral endovascular stent fractures compared with conventional energy-integrating detector CT (EID-CT), achieving up to 96% accuracy and identifying more than 80% of otherwise undetectable type 1 and 2 fractures, a phantom study showed.

METHODOLOGY

• Researchers placed five identical 7-mm balloon-expandable covered stents in an abdominal phantom and sequentially created four fracture types, ranging from single-strut fractures to complete transverse linear fracture with stent displacement.
• Stents were scanned using a PCD-CT system and a conventional EID-CT system under clinical acquisition protocols.
• PCD-CT images were reconstructed at ultra-high resolution (UHR; a slice thickness of 0.2 mm and a matrix size of 1024) and standard resolution (SR; a slice thickness of 0.4 mm and matrix sizes of 512 and 1024) using medium sharp and sharp kernels.
• Two independent radiologists with 5 and 7 years of experience in vascular imaging assessed diagnostic accuracy and image quality in a blinded randomised manner.
• Outcome measures included overall diagnostic accuracy; sensitivity; specificity; subjective image quality scores; and quantitative measurements of sharpness, blooming artefacts, and image noise.

TAKEAWAY

• PCD-CT with a sharp kernel and 1024 matrix achieved the highest overall diagnostic accuracy at 96% for UHR and 92% for SR, compared to PCD-CT with a 512 matrix and EID-CT (50%-56%).
• UHR, sharp-kernel PCD-CT detected 90% of cases of single-strut fractures and SR, sharp-kernel PCD-CT detected 80%, whereas PCD-CT with a 512 matrix and EID-CT failed to detect any single-strut fractures.
• PCD-CT with a sharp kernel and 1024 matrix improved sharpness (11,919 Hounsfield units [HU]/mm for UHR and 6779 HU/mm for SR vs 2318 HU/mm for EID-CT) and reduced blooming (0.612 and 0.645 mm vs 1.031 mm; P < .001 for all).
• Subjective quality scores for sharpness, blooming, and confidence reached medians of 5 for UHR and 4 for SR with sharp-kernel 1024-matrix PCD-CT, which were higher than those with EID-CT (P < .001 for all). 

IN PRACTICE

"When using a larger matrix and sharper reconstruction kernel, PCD-CT demonstrated significantly superior image quality compared to EID-CT, even at a reduced radiation dose," the authors wrote.

"These findings suggest that PCD-CT could enable earlier identification of stent instability, potentially reducing the risk of complications such as thrombosis and restenosis," they added.

SOURCE

The study was led by Yiheng Tan, University Medical Center Groningen, Groningen, Netherlands. It was published online on July 31, 2026, in the European Journal of Radiology.

LIMITATIONS

The study was limited by the ex vivo static phantom design, the absence of vessel pulsation and contrast medium, the evaluation of a single stent material and size, radiation dose differences across scanners, and the small sample size of five stents scanned once per CT system.

DISCLOSURES

No funding information was provided for this study. The authors declared having no competing interests.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.