Roundup of the Challenges in X-Ray Tomography workshop, ESRF, May 20-22 2015

Summary

This workshop focused on the following three topics, identified as recurrent challenges in the field of X-ray tomography:

“Large” objects exceeding the detector field of view (region-of-interest tomography)
Combined phase retrieval and tomographic reconstruction
Tomography from highly limited data: low-dose acquisition, continuous rotation, flat objects

A couple of months ahead of the event experimental dataset related to these topics were made available and a call for proposals was sent to the EXTREMA Cost action network, inviting mathematicians and algorithm developers to contribute with solution proposals.
The first day of the event was dedicated to beam line practicals and allowed the participants gaining insight into the process of data acquisition at a synchrotron source. Participants from the experiment community were offered a hands-on tutorial session on the open-source software library ASTRA, well suited for fast prototyping of algebraic, tomographic reconstruction algorithms.
During the oral sessions on the following days a number of promising strategies for addressing the reconstruction challenges were presented and discussed. Contributions concerning iterative phase retrieval algorithms, the “upgrade” of filtered backprojection with optimized filters and state of the art regularization techniques (dictionary learning, TV) retained special attention as they show clear potential to outperform and replace existing methods for a number of specific applications.

Presentations

Combined phase retrieval and tomographic reconstruction

Nghia Vo (Diamond Light Source, UK):Iterative X-ray phase contrast tomography using two measured intensities at flexible distance
Aike Ruhlandt (University of Gottingen, Germany): Tomography as a constraint for phase-retrieval
Max Langer (CREATIS, ESRF, France): Refractive index retrieval by combining the contrast transfer function and SART in X-ray in-line phase tomography
Paul Diemoz (UCL, UK): Methods for fast tomography in edge illumination X-ray phase-contrast imaging and other differential phase techniques
Federica Marone (PSI, Switzerland): Iterative 3D phase retrieval for ultrafast tomographic imaging

Region of interest tomography

Valery Titarenko (University of Manchester, UK): ROI tomography approaches used for Manchester-Diamond collaboration
Ville Kolehmainen (University of Eastern Finland): ROI tomography using wavelets
Daniel Pelt (CWI, The Netherlands): Approximating prior-based methods in a ROI

Limited data and low dose tomography

Alexey Buzmakov (Institute of Crystallography, Russia): Low dose and metal artifacts problems
Thomas Boulay (L2S, France): Bayesian 3D Computed Tomography
Pierre Paleo (ESRF, France): Compressed sensing reconstruction
Alberto Mittone (ESRF, France): Phase-retrieval methods in differential phase-contrast imaging
Jakob Joergensen (DTU, Denmark): Total variation regularized reconstruction from highly noisy projections

Files: