CT metal artefact reduction of total knee prostheses using angled gantry multiplanar reformation

This study was designed to determine whether or not acquiring CT images of total knee prostheses by using an angled gantry and multiplanar reformation can reduce beam hardening artefact. A CT phantom was created with a total knee prosthesis suspended in gelatine with a known attenuation. CT data was acquired with a gantry angled at 0 degrees, 5 degrees, 10 degrees and 15 degrees in both craniocaudal oblique planes. Axial images where then reformatted from these datasets. Two independent observers selected regions of interest to measure the mean and standard deviation (SD) of attenuation in the gelatine for all reformatted axial images. Artefact was measured as SD of the background attenuation and areas under the curve of SD for each gantry angle acquisition were compared. Inter-observer reliability was excellent (ICC=0.89, CI 0.875-0.908). The most accurate mean attenuation values for tissues around a TKR were obtained with a CT gantry using 10 degrees to 15 degrees anteroinferior to posterosuperior angulation. The smallest area under the curve for SD of attenuation for the whole prosthesis, and the femoral component in isolation, was obtained with a 5 degrees gantry angle in the same direction. The smallest area under the curve for the tibial component in isolation occurred with a gantry angle of 15 degrees. We conclude that acquiring CT data with a gantry angle can reduce metal artefact around a TKR. Optimal overall metal artefact reduction can be achieved with a small angle from anteroinferior to posterosuperior. Further selective artefact reduction around the tibial component can be achieved with larger angles.