Reduction of Metal Artefacts in Pediatric Imaging - New Possibilities Through Photon Counting Computed Tomography

Metal artefacts from medical implants significantly impair computed tomography (CT) image quality, necessitating their reduction. Photon counting technology helps to reduce artefacts through spectral information and monoenergetic reconstructions, but its use in children is limited due to radiation dose concerns. The aim of this study was to evaluate photon-counting-based metal artefact reduction (MAR) techniques for their applicability and diagnostic utility under dose-adapted conditions in pediatric patients.

A 10-year-old phantom with femoral nail osteosynthesis was studied on a photon-counting CT (NAEOTOM Alpha, Siemens Healthineers) using a pediatric reference protocol (70 kVp) without iterative metal artefact reduction (MAR), a 100Sn protocol with iterative MAR, Quantum-Plus (120 kVp) after dose adjustment (Computed Tomography Dose Index volume (CTDIvol) average: 1.30 mGy), and virtual monoenergetic reconstructions (VMI) from 60 to 190 keV. Image datasets were compared based on noise and artefact index (AIx) in different regions of interest (ROI).

MAR techniques reduced noise. Near osteosynthesis (ROI 1), noise reduction was 49.40% for the 100Sn protocol and 46.76% for the Quantum-Plus protocol. The artefact index was highest for the 70 kVp protocol at 58.73 and nearly halved for dose-adapted protocols with iterative MAR (100Sn: 32.90; Quantum-Plus: 29.83). Image noise decreased by 40.46% from 60 to 110 keV, with the greatest reduction near osteosynthesis (59.12). Artefact index was highest at 60 keV (18.54), decreased to 5.88 at 100 keV, and then increased to 7.58 at 190 keV.

All MAR techniques improved CT number accuracy and significantly reduced image noise, especially near the prothesis. Using the full spectral information protocol in pediatric imaging is reasonable after dose adjustment to pediatric reference values.