While the idea of randomly scrambling X-rays to extract information about a sample might seem counterintuitive, it has become integral to a highly sensitive X-ray multimodal imaging technique known as speckle-based imaging (SBI). Introduced a decade ago 1,2, this method has attracted growing interest in the X-ray imaging community. SBI delivers complementary absorption, phase-contrast, and dark-field (small-angle scattering) signals from a single dataset. It stands out among other phase-sensitive X-ray imaging methods due to its straightforward experimental setup, utilising readily available sandpaper as an optical element and eliminating the need for precise alignment or high stability of setup components. Various SBI data acquisition and analysis methods have been developed, with the Unified Modulated Pattern Analysis (UMPA) proving notably flexible and sensitive 3. These characteristics make SBI a promising tool in diverse fields such as materials science, industrial testing, and biomedicine, see example for virtual histology in Fig. 1. However, widespread adoption of SBI has been impeded by significant limitations in its ability to access shorter time and longer length scales: High-sensitivity, quantitative SBI, achieved through approaches like UMPA, is currently limited to long time scales, precluding dynamic studies, and to relatively short length scales, restricting the maximum sample size. The first limitation arises from the necessity to capture multiple images at various sandpaper positions for optimal image quality, while the second is due to the need for directly resolving the speckle pattern, requiring a small detector pixel size and resulting in a limited field of view. Here, we will present the status of our current efforts to address these limitations of SBI. Adapting the method to larger samples without compromising image quality requires decoupling the detector pixel size from measurement sensitivity. One approach is to incorporate an analyser mask in front of the detector, which translates the unresolved speckle pattern into measurable intensity changes in the detector pixels as the sandpaper or analyser is moved. While reminiscent of the G2 grating in X-ray grating interferometry, the mask does not require periodicity due to the random nature of the speckle pattern, and the scanning procedure differs. In the time domain, we aim to adapt UMPA for short exposure times, enabling the investigation of dynamic processes and evolving samples. As a first step, we will demonstrate stroboscopic SBI using UMPA, capturing a series of still images at regular intervals during a periodic or repeatable motion, which can later be combined to create a movie of the dynamic phenomenon. Multiple repetitions of the periodic process at different sandpaper positions allow for40retrospective sorting of images into process phases for each sandpaper step. Advancing SBI to accommodate larger samples and dynamic processes is expected to greatly boost its uptake for research, medical, and industrial applications, and extend its utility to new fields. A wide field of view will be indispensable for the integration of SBI into industrial manufacturing chains and clinical use. Applications of dynamic SBI may include in-situ biomechanical testing of bones and the visualisation of additive manufacturing processes. References 1 Morgan, K. S., et al., (2012). Appl. Phys. Lett., 100 (124102). 2 Berujon, S., et al., (2012). Phys. Rev. Lett., 108 (158102). 3 Zdora, M.-C., et al., (2017). Phys. Rev. Lett., 118 (203903). 4 Zdora, M.-C., et al., (2020). Optica, 7 (1221-1227).
Marie‐Christine Zdora (Mon,) studied this question.