MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel

Automated-Adaptive Iterative Metal Artifact Reduction: Quantitative Evaluation of Streak Artifact Reduction When Reconstructing CT Images in Presence of Metal

N Mistry1*, I Duba1 , A Halaweish1 , B Schmidt2 , C Hofmann2 , (1) Siemens Medical Solutions Inc. USA, Malvern, PENNSYLVANIA, (2) Siemens Healthcare GmbH, Erlangen

Presentations

(Wednesday, 8/1/2018) 10:15 AM - 12:15 PM

Room: Room 207

Purpose: Current metal artifact reduction (MAR) algorithms come in one of two flavors: a) a simple option independent of metal type or, b) an option that provides the user with a list of application-specific presets. However, both these implementations are suboptimal in the presence of multiple metal implants in a patient, for example a head & neck tumor patient with dental fillings and shoulder implants. Here we describe a new algorithm that aims to automate metal detection and artifact reduction, while providing different reconstructions at varying strengths of correction. We also aim to quantitatively evaluate the performance of the new algorithm in different clinical situations.

Methods: The automated-adaptive iterative metal artifact reduction (AiMAR) technique works slice-by-slice. MAR is carried out over multiple iterations based on user selected strength setting (value between 1 and 5) using frequency-split MAR and adaptive sinogram mixing. Beam hardening correction is applied for each type of metal. Performance of AiMAR is evaluated using total variation (TV) and a streak reduction ratio (SRR). Three different clinical cases were examined and the performance of the algorithm compared for the different strengths and against application-specific presets as well.

Results: In most clinical situations, it was observed that as the strength of the AiMAR algorithm increased the SRR increased as expected. A significant improvement was observed when comparing the performance of AiMAR and existing single preset selection for the same clinical indication. For example, the SRR using the different strengths increased 19.8% - 30.8%, as compared to the existing application preset that showed SRR at 12.6%.

Conclusion: AiMAR shows improved performance over existing MAR techniques by automating the metal detection and artifact reduction. Significant improvement is observed due to the inclusion of beam hardening correction for each metal type.

Funding Support, Disclosures, and Conflict of Interest: All authors are employed by either Siemens Medical Solutions Inc USA, or Siemens Healthcare GmbH

Keywords

Computed Radiography, Image Artifacts, Reconstruction

Taxonomy

IM- CT: Image Reconstruction

Contact Email