A continuous sampling scheme for edge illumination x-ray phase contrast imaging

Abstract

We discuss an alternative acquisition scheme for edge illumination (EI) x-ray phase contrast imaging (XPCi) based on a continuous scan of the object, and compare its performance to that of a previously used scheme, which involved scanning the object in discrete steps rather than continuously. By simulating signals for both continuous and discrete methods under realistic experimental conditions, the effect of the spatial sampling rate is analysed with respect to metrics such as image contrast and accuracy of the retrieved phase shift. Experimental results confirm the theoretical predictions.Despite being limited to a specific example, the results indicate that continuous schemes present advantageous features compared to discrete ones. Not only can they be used to speed up the acquisition, but they also prove superior in terms of accurate phase retrieval. The theory and experimental results provided in this study will guide the design of future EI experiments through the implementation of optimized acquisition schemes and sampling rates.

On the relative performance of edge illumination x‐ray phase‐contrast CT and conventional, attenuation‐based CT

“…In EI PCT, the PSF can be expressed as a multiplication of the source distribution (projected onto the sample plane) with a box function representing the presample mask aperture, which is then convolved with an additional box function representing the sample movement by distance d (sampling step) employed during the continuous dithering. 24,27 Notably, it is independent from the pixel size, provided the pixel response is approximately constant over the width of the detector mask aperture. Thus, if d is large compared to the presample mask aperture and the projected source, the PSF in EI PCT is approximately square with width d. In CT, this can be matched with a point source and pixels of size d with a perfectly square response.…”

“…In that case, d is equal to the sample displacement. Dithering can be performed in a step‐and‐shoot (the sample is kept stationary while the detector is acquiring, and is displaced during read‐out) or continuous manner (the sample is moved continuously throughout the acquisition) . In the following, it is assumed that EI PCT data are acquired with continuous dithering.…”

“…This allows neglecting factors relating to the detection efficiency in the subsequent formulas. Phase and attenuation tomograms are reconstructed with FBP with the Shepp–Logan filter. The spatial resolution in EI PCT and CT raw projections is the same, i.e., the setups have the same point spread function (PSF). In EI PCT, the PSF can be expressed as a multiplication of the source distribution (projected onto the sample plane) with a box function representing the presample mask aperture, which is then convolved with an additional box function representing the sample movement by distance d (sampling step) employed during the continuous dithering . Notably, it is independent from the pixel size, provided the pixel response is approximately constant over the width of the detector mask aperture.…”

Theoretical Framework for Spatial Resolution in Edge-Illumination X-Ray Tomography

“…A common way to achieve this is by means of a process called dithering, by which the sample is scanned laterally (along the x direction) in steps smaller than the sample-mask period, a frame is acquired at each step, and all frames are subsequently combined into a single, upsampled projection. Dithering can be performed either in step-and-shoot or continuous mode; in the former, the sample is displaced in between the acquisition of consecutive frames (i.e., while the detector is reading out), whereas in the latter the sample is displaced during the acquisition of frames (i.e., while the detector is integrating) [12]. While the effect of dithering on spatial resolution has previously been analyzed for two-dimensional (2D) (planar) edge-illumination imaging [10], an investigation for three-dimensional (3D) (tomographic) imaging, where sinograms are acquired rather than individual projections, is still lacking.…”

Asymmetric masks for laboratory-based X-ray phase-contrast imaging with edge illumination

“…We introduce an approach based on the use of an asymmetric mask that eliminates all movements typically associated with EI XPCI experiments, with exception of the sample which is scanned through the imaging system. Sample scanning also removes sampling problems that might be encountered when imaging an object through a small aperture 31 . The basic idea is that an asymmetric pattern of apertures and absorbing septa are designed in such a way that adjacent detector pixel columns receive different degrees of illumination.…”