Prototype metal artefact reduction algorithm in flat panel computed tomography - evaluation in patients undergoing transarterial hepatic radioembolisation.

ABSTRACT

OBJECTIVE: To investigate the effect of an on-site prototype metal artefact reduction (MAR) algorithm in cone-beam CT-catheter-arteriography (CBCT-CA) in patients undergoing transarterial radioembolisation (RE) of hepatic masses. METHODS AND MATERIALS Ethical board approved retrospective study of 29 patients (mean 63.7±13.7 years, 11 female), including 16 patients with arterial metallic coils, undergoing CBCT-CA (8s scan, 200 degrees rotation, 397 projections). Image reconstructions with and without prototype MAR algorithm were evaluated quantitatively (streak-artefact attenuation changes) and qualitatively (visibility of hepatic parenchyma and vessels) in near- (<1cm) and far-field (>3cm) of artefact sources (metallic coils and catheters). Quantitative and qualitative measurements of uncorrected and MAR corrected images and different artefact sources were compared RESULTS: Quantitative evaluation showed significant reduction of near- and far-field streak-artefacts with MAR for both artefact sources (p<0.001), while remaining stable for unaffected organs (all p>0.05). Inhomogeneities of attenuation values were significantly higher for metallic coils compared to catheters (p<0.001) and decreased significantly for both after MAR (p<0.001). Qualitative image scores were significantly improved after MAR (all p<0.003) with by trend higher artefact degrees for metallic coils compared to catheters. CONCLUSION: In patients undergoing CBCT-CA for transarterial RE, prototype MAR algorithm improves image quality in proximity of metallic coil and catheter artefacts. KEY POINTS • Metal objects cause artefacts in cone-beam computed tomography (CBCT) imaging. • These artefacts can be corrected by metal artefact reduction (MAR) algorithms. • Corrected images show significantly better visibility of nearby hepatic vessels and tissue. • Better visibility may facilitate image interpretation, save time and radiation exposure.