Quantitative analysis of orthopedic metal artefact reduction in 64-slice computed tomography scans in large head metal-on-metal total hip replacement, a phantom study

Abstract

Purpose: Quantification of the effect of O-MAR on decreasing metal artefacts caused by large head metal on metal total hip arthroplasty (MoM THA) in a dedicated phantom setup of the hip. Background: Pathological reactions of the hip capsule on Computed tomography (CT) can be difficult to diagnose due to different metal artefacts. The O-MAR algorithm deploys an iterative loop where the metal sinogram is identified, extracted, and subsequently serves as a mask to correct the measured sinogram. Main goal of this study is to quantify the ability of the O-MAR technique to correct deviation in medullary bone attenuation caused by streak artefacts from the large-head MoM THA embedded in a phantom. Secondary goal is to evaluate the influence of O-MAR on CNR. Methods: The phantom was designed as a Perspex box (PMMA) containing water and a supplementary MOM THA surrounded by Perspex columns comprising calibrated calcium pellets. Each column contains 200 mg of hydroxyapatite/calcium carbonate to simulate healthy bone tissue. Scans were obtained with and without a MoM THA at different dose levels. Different reconstructions were made with filter A, iDose4 level 5 and with and without O-MAR. The scans without the prosthesis were used as the baseline. Information about the attenuation in Hounsfield units, image noise in standard deviation within the ROI’s were extracted and the CNR was calculated. Results: Pellet L0 and R0 (proximal of the MoM THA) were defined as reference, lacking any disturbance by metal artefacts; L5, L6 and L8 were respectively visually categorized as ‘light’ ‘medium’ and ‘heavy disturbance’. Significant improvements in attenuation deviation caused by metal artefact were 43, 68 and 32 %, for respectively pellet L5, L6 and L8 (p < 0.001). Significant CNR improvements were present for L5 and L6 and were respectively 72 and 52 % (p < 0.001). O-MAR showed no improvement on CNR for L8. Conclusion: This phantom study significantly increases image quality by the use of O-MAR in the presence of metal artefacts by significantly reducing metal artefacts subsequently and increasing CNR on a 64 slice CT system in light and medium disturbance of the image.