Tomographic Image Reconstruction Using Compressed Sensing

Hiroyuki Kudo, Dong Jian, Katsumi Kamo, Noritaka Horii, Hiromitsu Furukawa, Satoshi Hata, Mitsuhiro Murayama, Kazuhisa Sato and Shinsuke Miyazaki

Abstract: In this paper, we review image reconstruction methods based on the Compressed Sensing (CS), which are attracting much attention in medical CT and electron tomography fields. The CS technique allows to reconstruct high-quality cross-sectional images from a small number of projection data, limited-angle projection data, and projection data with a low SN ratio. Furthermore, our recently developed CS-based algorithm, Iterative SEries Reduction (ISER), is overviewed. ISER was developed specifically for electron tomography applications.

Key words: Computed Tomography (CT), Electron Tomography, Image Reconstruction, Compressed Sensing, Total Variation