ABSTRACT
PURPOSE: To determine the value of a metal artefact reduction (MAR) algorithm with iterative reconstructions for dental hardware in carotid CT angiography. METHODS: Twenty-four patients (six of which were women; mean age 70 ± 12 years) with dental hardware undergoing carotid CT angiography were included. Datasets were reconstructed with filtered back projection (FBP) and using a MAR algorithm employing normalisation and an iterative frequency-split (IFS) approach. Three blinded, independent readers measured CT attenuation values and evaluated image quality and degrees of artefacts using axial images, multi-planar reformations (MPRs) and maximal intensity projections (MIP) of the carotid arteries. RESULTS: CT attenuation values of the internal carotid artery on images with metal artefacts were significantly higher in FBP (324 ± 104HU) datasets compared with those reconstructed with IFS (278 ± 114HU; P < 0.001) and with FBP on images without metal artefacts (293 ± 106HU; P = 0.006). Quality of IFS images was rated significantly higher on axial, MPR and MIP images (P < 0.05, each), and readers found significantly less artefacts impairing the diagnostic confidence of the internal carotid artery (P < 0.05, each). CONCLUSION: The MAR algorithm with the IFS approach allowed for a significant reduction of artefacts from dental hardware in carotid CT angiography, hereby increasing image quality and improving the accuracy of CT attenuation measurements. KEY POINTS: ⢠CT angiography of the neck has proven value for evaluating carotid disease ⢠Neck CT angiography images are often degraded by artefacts from dental implants ⢠A metal artefact reduction algorithm with iterative reconstruction reduces artefacts significantly ⢠Visualisation of the internal carotid artery is improved.