Research On CT Image Ring Artifact Reduction

CT technology has got fast development after the birth of the first clinical CT imaging system. Its scan range, rate and image quality has been improved hugely, and now CT is widely applied in clinical diagnosis, assistant therapy and other fields.It is by measuring the X-rays in all directions on the human body through the fault at the time of the cumulative attenuation coefficient (or projector), and then calculated by a computer the entire section on X-ray attenuation coefficient of the distribution, then the form is displayed in the image in order to help doctors diagnose the lesion. CT imaging can provide more than ordinary X-ray imaging of soft tissue higher resolution, and resolve the problem of overlapping three-dimensional structure of its medical imaging on epoch-making significanceArtifacts are resulted by virtual image which does not exist in the reconstructed images during scanning process. The artifacts are the impact of CT image quality, resulting in the loss of CT image diagnosis of one of the important factors. Ring artifacts in computed tomography (CT) images are formed by imperfect pixel channel in a detector, its existance degrade CT image, hamper post processing and prohibit quantitative analysis, due to it is similar between artifacts and Pathology image, it could lead to mis-diagnose. Therefore removing or at least significantly reducing such artifact is indispensable. Researching of the Ring Artifact and its related resolutions is one of the hot and key issues currently.Firstly this note systematically introduced the reasons of Ring Artifact, then a set of simple but efficient and fast processing method is proposed that effectively reduces ring artifacts in sinogram and already-reconstructed images. Correction for projection sinogram data is that the vertical line artifacts in the sinogram are removed by using edge function of Sobel operator to mark the edge of sinogram, and then using Hough or bwlabel function to find the position of ring artifact. At last, choose neighborhood interpolation, linear interpolation, cubic polynomial fitting interpolation to remove the ring artifact, and analyze and compare these three interpolation methods. Seeking the location of ring artifacts in Sine graph is the key step, but the final image quality after removal of artifacts mostly depends on the choosing of the way of the interpolation. Compared to other methods, the interpolation is the most simple and effective mothod.Correction for reconstructed images was that the ring artifacts in the Cartesian coordinate were transformed into the vertical line artifacts in the polar coordinate and removed as sinogram correction. The first method is a kind of image preprocessing. Another one belongs to image postprocessing. In most cases corrections are optimally applied to sinogram data. The set of algorithms enable virtually complete ring artifact removal from tomographic imagery with minimal to negligible contamination of the underlying data.