Combining monoenergetic extrapolations from dual-energy CT with iterative reconstructions: reduction of coil and clip artifacts from intracranial aneurysm therapy.

Abstract

To compare and to combine iterative metal artifact reduction (MAR) and virtual monoenergetic extrapolations (VMEs) from dual-energy computed tomography (DECT) for reducing metal artifacts from intracranial clips and coils. Fourteen clips and six coils were scanned in a phantom model with DECT at 100 and 150SnkVp. Four datasets were reconstructed: non-corrected images (filtered-back projection), iterative MAR, VME from DECT at 120keV, and combined iterative MAR + VME images. Artifact severity scores and visibility of simulated, contrast-filled, adjacent vessels were assessed qualitatively and quantitatively by two independent, blinded readers. Iterative MAR, VME, and combined iterative MAR + VME resulted in a significant reduction of qualitative (p < 0.001) and quantitative clip artifacts (p < 0.005) and improved the visibility of adjacent vessels (p < 0.05) compared to non-corrected images, with lowest artifact scores found in combined iterative MAR + VME images. Titanium clips demonstrated less artifacts than Phynox clips (p < 0.05), and artifact scores increased with clip size. Coil artifacts increased with coil size but were reducible when applying iterative MAR + VME compared to non-corrected images. However, no technique improved the severe artifacts from large, densely packed coils. Combining iterative MAR with VME allows for an improved metal artifact reduction from clips and smaller, loosely packed coils. Limited value was found for large and densely packed coils.