Proceedings of the IEEK Conference (대한전자공학회:학술대회논문집)

The Institute of Electronics and Information Engineers (대한전자공학회)
권호 표지

RECONSTRUCTION OF LIMITED-ANGLE CT IMAGES BY AN ADAPTIVE RESILIENT BACK-PROPAGATION ALGORITHM

  • Kazunori Matsuo (Department of Electrical & Electronics Engineering Faculty of Engineering, University of the Ryukyus) ;
  • Zensho Nakao (Department of Electrical & Electronics Engineering Faculty of Engineering, University of The Ryukyus) ;
  • Chen, Yen-Wei (Department of Electrical & Electronics Engineering Faculty of Engineering, University of The Ryukyus) ;
  • Fath El Alem F. Ah (Department of Management and Information Systems Meio University)
  • Published : 2000.07.01
Download PDF
⟨ Previous Next ⟩

Abstract

A new and modified neural network model Is proposed for CT image reconstruction from four projection directions only. The model uses the Resilient Back-Propagation (Rprop) algorithm, which is derived from the original Back-Propagation, for adaptation of its weights. In addition to the error in projection directions of the image being reconstructed, the proposed network makes use of errors in pixels between an image which passed the median filter and the reconstructed one. Improved reconstruction was obtained, and the proposed method was found to be very effective in CT image reconstruction when the given number of projection directions is very limited.

Keywords