Related Experiment Video
Updated: Jun 20, 2026
Assessment of Bone Fracture Healing Using Micro-Computed Tomography
Published on: December 9, 2022
Photon-counting CT: image quality evaluation in patients with tibial plateau fracture treated with metallic
Ann-Sofi Björkman1,2,3, Alexandr Malusek4,5, Anders Persson4,6,5
1Department of Health, Medicine and Caring Sciences, Linköping University, Linköping, Sweden. ann-sofi.bjorkman@liu.se.
Objective:
The aim was to compare the image quality of photon-counting detector CT (PCD-CT) and energy-integrating detector CT (EID-CT) in patients with tibial plateau fractures treated with metallic osteosynthesis material, and to identify optimal reconstruction parameters for PCD-CT.
Materials And Methods:
After ethical approval, twelve patients underwent PCD-CT and EID-CT scans. Images were reconstructed using bone and soft-tissue kernels with metal artifact reduction (iMAR). PCD-CT virtual monoenergetic images (VMI) at 70, 110, and 150 keV were generated. Five radiologists assessed metal artifact severity and bone and soft-tissue visualization using a 7-point Likert scale. Visual grading characteristics analysis was performed. Noise levels were quantified and compared using Wilcoxon's signed-rank test.
Results:
EID-CT was rated superior in reducing metal-artifact streaks (AUC: 0.10-0.21). No significant difference was found between EID-CT iMAR and PCD-CT VMI at 110 keV and 150 keV (AUC: 0.40-0.49) concerning the metal-bone interface. An ultra-high-resolution PCD-CT kernel outperformed its EID-CT counterpart in bone visualization, with AUC values of 0.67-0.92 across all bone criteria, including those incorporating artifact-affected regions. PCD-CT showed lower noise.
Conclusion:
Metal artifact reduction was superior in EID-CT iMAR images compared to PCD-CT iMAR and VMI 110/150 keV, except at the metal-bone interface, where EID-CT iMAR images and PCD-CT VMI 110/150 keV performed comparably. The ultra-high-resolution PCD-CT kernel provided the best bone visualization, even when artifact-affected areas were included. Noise levels were lower in PCD-CT. This is the first in vivo comparison of photon-counting and energy-integrating CT for postoperative knee imaging with metallic osteosynthesis material.
Relevance Statement:
These findings highlight the need for improved metal artifact reduction in PCD‑CT, while demonstrating superior bone visualization, and support a complementary interpretation strategy in postoperative knees using high‑energy VMI and MAR‑corrected reconstructions.
Key Points:
Photon-counting CT demonstrated excellent bone visualization. Conventional EID-CT achieved better metal artifact reduction than photon-counting CT. The metal-bone interface was rated similarly for energy-integrating CT and photon-counting CT virtual monoenergetic images. These findings indicate that further improvements in metal artifact reduction algorithms for PCD-CT are warranted.
