Research Of Metal Artifacts Reduction In Medical X-CT Image

Nowadays X-ray computed tomography has been one of the most important diagnostic tools in medicine. After the x-ray attenuation coefficients through a cross-section of an object are measuraed in different directions, the distribution of the x-ray attenuation in this cross-section is calculated by a computer, and then it will be diplayed to help doctors to detect the position of a focus. The attenuation coefficients of the focus are always different from the normal conditions, so the distribution of x-ray attenuation coefficient can provide important information for focus detection.However, an important problem with x-ray computed tomography is metal artifacts reduction. Because of the higher atomic number, metal objects's attenuatation to the diagnostic x-rays are much greater than soft tissues and bones, and much fewer photons can reach detectors. It is well known that when the filtered back-projection (FBP) method is used to reconstruct image, the projections must be complete. So when a metal is present, pronounced dark and bright streaks are produced in the reconstruction with FBP. These artifacts seriously degrade image quality particularly near the metal areas; therefore they will make detection very difficult or even impossible.This paper starts from the basic theory of CT image reconstruction and the methods of metal artifact reduction are researched in two aspects. Firstly from the point of projection compensation, considering the projections are missed or distorted due to the presence of metal object, the method of projection compensation based on linear interpolation is used to compensate the missing data before reconstruction. Secondly, from the point of reconstruction method, the FBP method is not proper again because of the incompleteness of projection, so the iterative reconstruction methods are used here, while the projections of metal are considered missing and are not used in the iterative process. The iterative reconstruction methods include the algebraic reconstruction technique (ART) and the other twos—MLEM and OSEM based on statistical iterative. Images can also be reconstructed by them satisfactorily. The results of numerical simulation experiments show that the metal artifacts caused by the metallic objects can be suppressed to a great extent by the proposed methods.