Optimization of Post-processing Reconstruction for Orthopedic Metal Implants on Photon-Counting Detector CT: A Within-Patient Comparison

Abstract Objectives To determine how iterative metal artifact reduction (iMAR), virtual monoenergetic imaging (VMI), and reconstruction kernel selection affect metal artifacts and image quality around orthopedic implants on photon-counting detector CT (PCD-CT), and to identify task-specific reconstructions for bone and soft-tissue evaluation. Materials and methods This retrospective within-patient study included 10 patients with 11 orthopedic implant sites. All reconstructions were generated from the same PCD-CT raw data. Three predefined comparisons were performed: iMAR off versus on at 70 keV/Br40/1.0 mm; four bone reconstructions combining T3D or 145-keV VMI with Qr64 or Br64 at 0.4 mm; and four soft-tissue reconstructions combining T3D or 65-keV VMI with Br40 or Qr40 at 1.0 mm. CT attenuation, image noise, artifact index (AI), signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), 5-point Likert scores, and interreader agreement were evaluated. Results With matched parameters, iMAR increased median bone scores from 2 to 4 and soft-tissue scores from 1 to 2–3. For bone, 145-keV VMI reduced image noise and AI compared with T3D; 145-keV Br64 had the highest subjective scores, whereas 145-keV Qr64 showed the most favorable quantitative metrics. For soft tissue, T3D had lower noise and AI and higher SNR than 65-keV VMI, and T3D Br40 received the highest subjective scores. Weighted κ ranged from 0.635 to 0.717. Conclusion Post-processing optimization on PCD-CT was task specific. iMAR improved image evaluability, 145-keV VMI favored bone and bone-implant interface assessment, and T3D Br40 favored adjacent soft-tissue evaluation.