Slice-based visualization is a common analysis tool for 3D tomographic reconstructions. Conventionally, this requires a full volume, whereas a quasi-3D approach is to only reconstruct those slices that are currently of interest. Direct algorithms such as FBP and FDK are local in that they can reconstruct arbitrary slices efficiently. In contrast, iterative algorithms are harder to restrict to slices. We discuss approaches to tackle this problem, which is closely related to region of interest tomography.

This presentation is part of Minisymposium “MS33 - Advances in reconstruction algorithms for computed tomography (4 parts)”
organized by: Gunay Dogan (Theiss Research, NIST) , Harbir Antil (George Mason University) , Elena Loli Piccolomini (Dept. Computer Science and Engineering, University of Bologna) , Samuli Siltanen (University of Helsinki) .

Authors:

Holger Kohr (Thermo Fisher Scientific)

Jan-Willem Buurlage (Centrum Wiskunde & Informatica)

Joost Batenburg (CWI, Amsterdam)

Willem Jan Palenstijn (Centrum Wiskunde & Informatica, Amsterdam)

Keywords:

computer vision, image reconstruction, inverse problems, numerical linear algebra