In multispectral (or photon-counting) computed tomography (MCT), accurate image reconstruction remains challenging because the appropriate data model in MCT is highly non-linear. In the presentation, our recent work will be reported on non-convex optimization-based image reconstruction (OBIR) approach to directly solving adequately the non-linear data model in MCT. Following the derivation of the OBIR approach, its effectiveness will first be demonstrated for image reconstruction in the standard MCT. Subsequently, a study will be carried out to reveal that the OBIR approach can be exploited for enabling MCT with partial scanning configurations yet involving no or minimum hardware modification to standard CT, thus lowering hardware cost, enhancing scanning flexibility, and reducing imaging dose/time. Numerical studies are carried out for demonstration of enabled image reconstruction for partial scanning configurations of practical significance with varying scanning angular range and/or x-ray illumination coverage in MCT imaging.

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:

Xiaochuan Pan (University of Chicago)

Keywords:

image reconstruction, inverse problems, nonlinear optimization