Reducing metal artefacts and radiation dose in musculoskeletal CT imaging

In computed tomography (CT) metal hardware such as total hip arthroplasties (THA) and intramedullary nails and plates used for osteosynthetic fixation cause metal artefacts with different severities due to differences in size, shape, geometry and density. Modifying image acquisition, image reconstruction, projection data and/or image data and use of virtual monochromatic imaging extracted from dual-energy CT all contribute to the reduction of metal artefacts. In this thesis we focused on the clinical value of novel CT techniques including model-based iterative reconstruction, metal artefact reduction software and dual-energy CT in reducing metal artefacts and radiation dose in musculoskeletal imaging. Metal artefact reduction, improvement of overall image quality and the effects of radiation dose reduction were investigated in phantom, human cadaveric and patient studies based on quantitative and qualitative outcome measures. The clinical value and limitations of these novel techniques separately and combined were evaluated in order to tailor these techniques for patient groups and indications.
Results showed that the effectiveness of metal artefact reduction (MAR) depends on the used MAR approach and hardware type. Furthermore, using model-based iterative reconstruction combined with aggressive radiation dose reduction in patients with THA is discouraged, especially when evaluating osseous structures and when focussing on small details. Providing detailed information regarding the composition and size of the hardware is important. In this way the imaging chain, from acquisition to reconstruction, post-processing and visualisation can be tailored to reduce metal artefacts and radiation dose, improve image quality and improve the diagnostic value of CT in musculoskeletal radiology.