Observer Evaluation of a Metal Artifact Reduction Algorithm Applied to Head and Neck Cone Beam Computed Tomography Images

Abstract

To quantify, through an observer study, the reduction in metal artifacts on cone beam computed tomography (CBCT) images using a projection-interpolation algorithm, on images containing metal artifacts from dental fillings. A projection-interpolation algorithm was applied to CBCT images containing metal artifacts from dental fillings. This technique involved identifying metal regions in individual CBCT projections and interpolating the surrounding values to remove the metal from the projection data. Axial CBCT images were then reconstructed, resulting in a reduction in the streak artifacts produced by the metal. Five radiation oncologists were subsequently asked to rank images (before and after metal artifact reduction) using a modified 5-point Likert scale. The scale consisted of the following: 1, very poor image quality, severe artifacts; 2, poor image quality, major artifacts; 3, moderate image quality, some artifacts; 4, good image quality, limited artifacts; and 5, excellent image quality, no artifacts. The interpolation substitution technique successfully reduced metal artifacts in all cases. From a total of 28 images, the mean Likert score on the original images was 2.0 +/- 0.8, while the mean score on the postartifact reduction images was 3.3 +/- 0.7 (P<0.0001). The mean difference in ranking score between uncorrected and corrected images was 1.4 with a 95% confidence interval (1.2, 1.6). Corrected images had fewer or no streak artifacts compared to their uncorrected counterparts. Artifact correction tended to perform better on patient images with less complex metal objects versus those with multiple large dental fillings. The interpolation substitution is an efficient and effective technique for reducing metal artifacts caused by dental fillings on CBCT images as demonstrated by the statistically significant increase in observer image quality ranking score between corrected and uncorrected images. This technique may be effective in reducing such artifacts in patients with head and neck cancer receiving daily image guided radiation therapy. The long-term goals of this study are to develop a user interface for application of this technique in head and neck image guided radiation therapy, thereby reducing metal artifacts and improving interpretation of CBCT image data.