Superiorized iteration algorithm for ct image simultaneous reconstruction and segmentation
INVERSE PROBLEMS AND IMAGING(2024)
摘要
. We propose a model for computed tomography (CT) image reconstruction and segmentation simultaneously, which can be applied to singleenergy spectral (traditional) CT (SECT) and dual-energy spectral CT (DECT) image reconstruction. It is difficult to solve the model due to the large scale of imaging system and the nonlinear forward projection of DECT. In order to deal with the model, a superiorized iteration algorithm is presented, which handles image segmentation and image reconstruction alternately. The two steps are combined by a superiorized perturbation step. The convergence of the iteration procedure is proved for SECT image reconstruction theoretically, and the convergence for DECT image reconstruction is investigated numerically. Experiments on various data are conducted. Comparisons with existing methods show that the proposed method is better quantitatively and visually.
更多查看译文
关键词
CT image reconstruction,spectral CT,image segmentation,superiorized iteration,piecewise constant
AI 理解论文
溯源树
样例
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要