Application of a Novel Metal Artifact Correction Algorithm in Flat-Panel CT After Coil Embolization of Brain Aneurysms: Intraindividual Comparison

PURPOSE To evaluate a novel algorithm for correcting beam hardening artifacts caused by metal implants in computed tomography performed on a C-arm angiography system equipped with a flat panel (FP-CT). MATERIALS AND METHODS 16 datasets of cerebral FP-CT acquisitions after coil embolization of brain aneurysms in the context of acute subarachnoid hemorrhage have been reconstructed by applying a soft tissue kernel with and without a novel reconstruction filter for metal artifact correction. Image reading was performed in multiplanar reformations (MPR) in average mode on a dedicated radiological workplace in comparison to the preinterventional native multisection CT (MS-CT) scan serving as the anatomic gold standard. Two independent radiologists performed image scoring following a defined scale in direct comparison of the image data with and without artifact correction. For statistical analysis, a random intercept model was calculated. RESULTS The inter-rater agreement was very high (ICC = 86.3 %). The soft tissue image quality and visualization of the CSF spaces at the level of the implants was substantially improved. The additional metal artifact correction algorithm did not induce impairment of the subjective image quality in any other brain regions. CONCLUSION Adding metal artifact correction to FP-CT in an acute postinterventional setting helps to visualize the close vicinity of the aneurysm at a generally consistent image quality.