Simultaneous Reconstruction and Segmentation of CT Scans with Shadowed Data

François Lauze (1) , Yvain Quéau (2) , Esben Plenge (1)

1 DIKU - Department of Computer Science [Copenhagen]
2 TUM - Technische Universität Munchen - Université Technique de Munich [Munich, Allemagne]

We propose a variational approach for simultaneous reconstruction and multiclass segmentation of X-ray CT images, with limited field of view and missing data. We propose a simple energy minimi-sation approach, loosely based on a Bayesian rationale. The resulting non convex problem is solved by alternating reconstruction steps using an iterated relaxed proximal gradient, and a proximal approach for the segmentation. Preliminary results on synthetic data demonstrate the potential of the approach for synchrotron imaging applications.

Domains: Computer Science [cs]; Computer Vision and Pattern Recognition [cs.CV]

François Lauze, Yvain Quéau, Esben Plenge. Simultaneous Reconstruction and Segmentation of CT Scans with Shadowed Data. Sixth International Conference on Scale Space and Variational Methods in Computer Vision (SSVM), Jun 2017, Kolding, Denmark. pp.308-319, ⟨10.1007/978-3-319-58771-4_25⟩. ⟨hal-02118564⟩