Dirk Wagenaar, Abstract Co-Author: Nothing to Disclose

Emiel Van der Graaf, Abstract Co-Author: Nothing to Disclose

Marcel Greuter PhD, Presenter: Nothing to Disclose

Typical streak artefacts known as metal artefacts occur in the presence of strongly attenuating materials in computed tomography (CT). Although previous studies successfully compared different metal artifact reduction (MAR) methods, a comparison of commercially available MAR methods is yet to be made. The aim of this study was to quantify the difference in effectiveness of the currently available MAR methods of different vendors.

Three vendors (A, B and C) were asked to use their preferential CT scanner for applying their MAR methods in a head and neck CT scan. The scans were made using an anthropomorphic head and neck phantom (Kyoto Kagaku, Japan). Three dental implants were constructed using dental grade amalgam which were inserted between the phantom’s teeth. Three CT scans were made with amalgam implants and three without amalgam implants. The three scans made without metal implants were averaged to create a single reference image for each vendor to which the other reconstructions were compared. The acquisition and reconstruction parameters were defined by the vendor for optimal MAR, using a CTDIvol of 9 mGy. Registration software (Mirada Medical, Denver, USA) was used to compensate for the movement of the phantom between the scans. The root mean square error due to MA’s (RMSEMA) was calculated for all reconstructions in the proximity of the amalgam implants.

For all vendors the reconstructions made without MAR showed a similar RMSEMA of 110.9±0.3, 97.5±10.0 and 107.5±4.2 HU for vendor A, B and C respectively. The reconstructions made with MAR showed a RMSEMA reduction of 42±1%, 38±8% and 18±6% reduction for vendor A, B and C, respectively. The difference in RMSEMA was -6% and +37% between the MAR methods of vendors A and B and vendors A and C, respectively.

For reconstructions made with each vendors best performing MAR method the RMSEMA of reconstructions made by vendor A and C were 7% and 47% higher than the RMSEMA of reconstructions made by vendor B.

Efficiency of metal artifact reduction algorithms differ in computed tomography systems from different vendors, which may have an impact on the clinical outcome.

Wagenaar, D, Van der Graaf, E, Greuter, M, Quantitative Comparison of Metal Artifact Reduction Methods in Computed Tomography by Different Vendors.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14006588.html