Iterative approach to partial-volume artifact reduction in CT

摘要

Abstract: One of the basic assumptions in the computed tomography(CT) is that the scanned object has constantattenuation characteristics across the thickness of theslice. In reality, however, this assumption is oftenviolated. The projection data set of a head CT scan,for example, is often corrupted by bony structurespartially intruded into the scanning plane. As aresult, severe streaking and shading artifacts appearin the reconstructed images. This phenomenon is calledpartial volume. In this paper, we propose an iterativeapproach to the partial volume artifact reduction. CTimages are first reconstructed with a filteredbackprojection algorithm. The generated imagessubsequently undergo a fuzzy membership classificationprocess to arrive at bone-only images, which in turnwill be used to produce gradient images. The projectionerror is then calculated based on the gradient image.For better error estimation, the scan data is collectedin a helical mode and highly overlapped images arereconstructed. The error term is filtered andback-projected to produce a partial volume error image,which is scaled and subtracted from the original image.Various phantom studies have demonstrated theeffectiveness of our approach.!6