Detector requirements for single mask edge illumination x-ray phase contrast imaging applications

TL;DR

This paper discusses simplified detector setups for edge illumination X-ray phase contrast imaging, enabling single-shot multi-modal imaging to reduce acquisition time and errors, thus broadening application potential.

Contribution

It introduces two novel simplified set-ups leveraging advanced detector technology for single-shot multi-modal XPCI, improving efficiency and accuracy over traditional methods.

Findings

Achieved single-shot multi-modal imaging with simplified setups.

Reduced image acquisition time and errors in XPCI.

Enhanced potential for challenging imaging applications.

Abstract

Edge illumination (EI) is a non-interferometric X-ray phase contrast imaging (XPCI) method that has been successfully implemented with conventional polychromatic sources, thanks to its relaxed coherence requirements. Like other XPCI methods, EI enables the retrieval of absorption, refraction and ultra-small angle X-ray scattering (USAXS) signals. However, current retrieval algorithms require three input frames, which have so far been acquired under as many different illumination conditions, in separate exposures. These illumination conditions can be achieved by deliberately misaligning the set-up in different ways. Each one of these misaligned configurations can then be used to record frames containing a mixture of the absorption, refraction and scattering signals. However, this acquisition scheme involves lengthy exposure times, which can also introduce errors to the retrieved signals. Such errors have, so far, been mitigated by careful image acquisition and analysis. However, further reduction to image acquisition time and errors due to sample mask/sample movement can increase the advantages offered by the EI technique, and enable targeting more challenging applications. In this paper, we describe two simplified set-ups that exploit state-of-the-art detector technologies to achieve single-shot multi-modal imaging.