Sphoorti Shellikeri, Presenter: Research funded, Siemens AG

Erin Girard PhD, Abstract Co-Author: Employee, Siemens Corporation

Anne Marie Cahill MBBCh, Abstract Co-Author: Nothing to Disclose

Metallic objects in interventional procedures can induce artifacts in C-arm CT images, reducing the image quality and potentially affecting diagnostic decisions. The purpose of this study is to evaluate the efficacy of a Metal Artifact Reduction (MAR) algorithm in reducing metal artifacts in C-arm CT images acquired during pediatric bone biopsy procedures.

C-arm CT images (syngo DynaCT, Artis Zee VC14, Siemens AG, Forchheim, Germany) were acquired during 8 bone biopsy procedures using an 8s-DR low dose protocol (0.1/0.17 µGy/p entrance dose). Two C-arm CT images of the biopsy needle were also acquired in an anthropomorphic torso phantom. All images were reconstructed offline on a clinical workstation (syngo XWP, VB15D) and corrected on a research workstation using the MAR prototype. The uncorrected and corrected images were qualitatively assessed by 2 readers of varying experience using a 3-point Likert scale (2: no/minimal artifact; 1: moderate artifact; 0: massive artifact). Additionally, the images were assessed quantitatively using mean and standard deviation, in Hounsfield unit, measured in 6 regions of interest in each data set (4 ROI's surrounding the needle and 2 distant from the needle).

8/10 images corrected using MAR algorithm scored a 2 on the Likert scale, suggesting an improvement in the image quality with no/minimal artifacts and a very good needle and soft tissue visualization. There was no change in image quality for 2/10 images which already had no/minimal artifacts. The MAR algorithm showed significant reduction in metal artifacts in regions surrounding the needle with noise reduction up to 177.25 HU (avg. reduction of 89.41 HU) in the patient group, and up to 289 HU (avg. reduction of 135.88 HU) in the phantom group. In tissue and air regions distant from the needle, noise level remained relatively unchanged (up to 7.5 HU noise reduction in the patient group and 2.5 HU noise increase in the phantom group).

Our initial experience with this C-arm CT MAR algorithm shows significant reduction in metal artifacts and improved image quality around the needle and the soft tissue in C-arm CT images acquired during pediatric bone biopsy procedures.

The MAR algorithm can be used to reduce metal artifacts in C-arm CT images, acquired during interventional pediatric bone biopsy procedures, for better visualization of the needle and soft tissue.

Shellikeri, S, Girard, E, Cahill, A, Initial Experience with Metal Artifact Reduction (MAR) Algorithm on C-arm CT Images Acquired during Pediatric Bone Biopsy Procedures.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14008208.html