In case of the presence of metallic implants in patients, image quality in computed tomography (CT) often suffers from severe artifacts which may cause the image to be non-diagnostic. Traditional sinogram interpolation based metal artifact reduction (MAR) algorithms replace metal-corrupted projection data (the metal trace) via various interpolation strategies. However, while these methods can effectively remove metal artifacts, new residual artifacts will often emerge as surrogate data cannot be completely accurate. The secondary artifacts are usually as unacceptable as original metal artifacts. Therefore, better projection data estimation is critical. In this work, the authors proposed a prior-interpolation method for a better estimation of the metal-corrupted projection data. The proposed method requires an intermediate image, called the prior image and the prior image is generated via an image post-processing strategy, containing pre-filtering, bone extraction and soft-tissue restoring. This prior image is then forward projected to guide the data correction of the metal trace using a smooth sinogram completion technique. The corrected projections are used to reconstruct the final corrected image. The proposed algorithm is tested on hip and head CT test images containing metal objects. A comparison with other existing metal artifact reduction methods demonstrates that the prior-interpolation method performs better in residual artifacts suppression and tissue feature preservation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.