A novel noniterative metal artifact reduction method using coherence transport with fast marching in computed tomography

  • Authors:

  • Yi Zhang;Yi-Fei Pu;Wei-Hua Zhang;Jin-Rong Hu;Han Wang;Yin-Jie Lei;Ji-Liu Zhou

  • Affiliations:

  • College of Computer Science, Sichuan University, Chengdu, China;College of Computer Science, Sichuan University, Chengdu, China;College of Computer Science, Sichuan University, Chengdu, China;School of Mathematics and Computer Engineering, Xihua University, Chengdu, China;Department of Physics, Ecole Normale Super Paris, Paris, France;School of Computer Science and Software Engineering, The University of Western Australia, Crawley, Australia;College of Computer Science, Sichuan University, Chengdu, China

  • Venue:
  • International Journal of Imaging Systems and Technology
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Metal artifact reduction (MAR) is still a challenging problem in current computed tomography reconstruction. Traditional methods usually use the unaffected projection data in the same degree to interpolate contaminated positions in the bins. The continuality and smoothness features become hard to fulfill. In this article, we propose a novel noniterative MAR method that uses coherence transport with fast marching. Two-dimensional information, including the neighboring bins, is used to inpaint the uncorrected sinogram with a special form of weighted average according to a sequence calculated by the fast marching method. The results of the experiments and the quantitative analysis demonstrate that the proposed method can effectively suppress the artifacts caused by metal with a high-attenuation coefficient. Compared with linear interpolation, the proposed method can introduce less streak artifacts and obtain more accurate structure information of organs. © 2012 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 22, 200–207, 2012 © 2012 Wiley Periodicals, Inc.