Low-Energy X-ray Fluorescence (XRF) mapping at synchrotrons is limited by acquisition time and dose, especially for sensitive samples. Sparse scanning and traditional denoising methods may struggle with high noise and low photon counts. While supervised deep learning can hallucinate “nice looking” features when ground truth is unavailable, Deep Image Prior (DIP) methods avoid pre-training by optimising a network on a single measured image. In this work, we extend DIP with Bayesian reasoning and a Poisson likelihood tailored to XRF maps. We test our method on noisy synchrotron XRF data, demonstrating reliable image reconstruction.

Abstract Computed Tomography (CT) is fundamentally an inverse problem combining linear operators, regularization and discrete inference. Artificial intelligence has improved reconstruction, denoising and segmentation. Quantum Computing (QC) is typically discussed in terms of computational speed. For CT, the more relevant question is structural compatibility. Several CT subproblems admit formulations that are aligned with quantum-native primitives for structured linear algebra and quadratic optimization. This short paper identifies three directions: reconstruction, denoising and segmentation. It briefly formalizes each component and details a QUBO-based formulation for the segmentation stage implemented in a prototype demonstrator, QUBOSegment . We also report ongoing development of a production-oriented system, NextGenSegment (NGS) , built on our Modular Adaptive Processing Infrastructure (MAPI). This work is intentionally scoped as a formulation- and workflow-oriented proof of concept rather than a benchmarking study claiming quantum performance advantage. The aim is to expose the CT community to technically grounded QC formulations and to encourage systematic benchmarking in realistic synchrotron and laboratory settings.

PERCIVAL is a novel soft X-ray detection system designed for the needs of modern microscopy. By integrating it into the TwinMic end-station at Elettra Sincrotrone Trieste, we conducted an exploratory computational microscopy experiment on biological samples, aiming at evaluating the entire system in a real use-case scenario. We present the methodology to convert the RAW data and our high-resolution image reconstructions.

X-ray fluorescence (XRF) spectroscopy is a versatile tool employed in various fields like biomedicine, nanotechnology, and cultural heritage, providing comprehensive elemental maps that combine spectrographic, spatial, and morphological data. While commonly utilized in laboratories and synchrotron beamlines, accurate XRF data interpretation demands specialized software that incorporates the underlying physics, with PyMCA (ESRF) being a notable example. This manuscript introduces XRFitVis, an innovative, cloud-based application designed for the effective visualization of mu XRF maps, akin to those produced by PyMCA and similar platforms. This modern tool offers a user-friendly, web-based solution that facilitates the assessment of experimental results and the creation of publication-ready figures, eliminating the need for local installation. Garnering positive feedback from early adopters, XRFitVis is now available to a broader audience. Hosted by Elettra Sincrotrone Trieste primarily for PyMCA-fitted results, it aims for broad compatibility with various XRF fitting software, marking a significant step forward as the first free, generic, yet XRF-specific visualization tool, poised to benefit the global spectroscopy community.

Numerical simulation algorithms for optics design, computational imaging and wave-matter interaction are often implemented in legacy or non-free languages such as Fortran77/90, C/C++ or Matlab. Even if effective, these old codebases may pose challenges for hardware acceleration and parallelization of existing algorithms, two key aspects of modern digital twinning. In this paper, we propose to use Large Language Models (LLMs) to assist in the development, translation and optimization of such algorithms. By testing different free off-the-shelf alternatives, we evaluate our approach on several numerical simulation tasks in computational microscopy.

Computational techniques allow breaking the limits of traditional imaging methods, such as time restrictions, resolution, and optics flaws. While simple computational methods can be enough for highly controlled microscope setups or just for previews, an increased level of complexity is instead required for advanced setups, acquisition modalities or where uncertainty is high; the need for complex computational methods clashes with rapid design and execution. In all these cases, Automatic Differentiation, one of the subtopics of Artificial Intelligence, may offer a functional solution, but only if a GPU implementation is available. In this paper, we show how a framework built to solve just one optimisation problem can be employed for many different X-ray imaging inverse problems.

The possibility to obtain a three-dimensional representation of a single object with sub-μm resolution is crucial in many fields, from material science to clinical diagnostics. This is typically achieved through tomography, which combines multiple two-dimensional images of the same object captured at different orientations. However, this serial imaging method prevents single-shot acquisition in imaging experiments with free electron lasers. In the present experiment, we report on a new approach to 3D imaging using extreme-ultraviolet radiation. In this method, two EUV pulses hit simultaneously an isolated 3D object from different sides, generating independent coherent diffraction patterns, resulting in two distinct bidimensional views obtained via phase retrieval. These views are then used to obtain a 3D reconstruction using a ray tracing algorithm. This EUV stereoscopic imaging approach, similar to the natural process of binocular vision, provides sub-μm spatial resolution and single-shot capability. Moreover, ultrafast time resolution and spectroscopy can be readily implemented, and a further extension to X-ray wavelengths can be envisioned as well.

Scanning microscopies and spectroscopies like X-ray Fluorescence (XRF), Scanning Transmission X-ray Microscopy (STXM), and Ptychography are of very high scientific importance as they can be employed in several research fields. Methodology and technology advances aim at analysing larger samples at better resolutions, improved sensitivities and higher acquisition speeds. The frontiers of those advances are in detectors, radiation sources, motors, but also in acquisition and analysis software together with general methodology improvements. We have recently introduced and fully implemented an intelligent scanning methodology based on compressive sensing, on a soft X-ray microscopy beamline. This demonstrated sparse low energy XRF scanning of dynamically chosen regions of interest in combination with STXM, yielding spectroimaging data in the megapixel-range and in shorter timeframes than were previously not feasible. This research has been further developed and has been applied to scientific applications in biology. The developments are mostly in the dynamic triggering decisional mechanism in order to incorporate modern Machine Learning (ML) but also in the suitable integration of the method in the control system, making it available for other beamlines and imaging techniques. On the applications front, the method was previously successfully used on different samples, from lung and ovarian human tissues to plant root sections. This manuscript introduces the latest methodology advances and demonstrates their applications in life and environmental sciences. Lastly, it highlights the auxiliary development of a mobile application, designed to assist the user in the selection of specific regions of interest in an easy way.

X-ray fluorescence (XRF) spectroscopy is a widely used technique in microscopy, spanning from biology to cultural heritage applications. Its purpose is to characterize qualitatively and quantitatively, the presence of elemental species in a sample. This is accomplished through fitting the acquired data to a Gaussian model, identifying which XRF lines and associated elements are present. As a result, 2D images of cumulative count-rate maps associated with each element are produced. This procedure is not trivial to apply efficiently in a workflow, as it requires the user to be able to set a series of parameters (e.g., beam energy, background subtraction, etc.) on top of selecting XRF lines under investigation. Furthermore, users should easily and swiftly be able to change setup parameters and evaluate the effects on the results. In the present work, we introduce a web-based application that allows users to load the XRF data, setup a fit and inspect the results interactively within a simple graphical user interface (GUI) that enables easily going back and forth from setup to result inspection. In particular, it is possible to quickly view the count-rate maps and curve fitting simultaneously, on any single pixel spectra present in the images. The web-application can be accessed locally by a web-browser, but runs remotely on a cloud, freeing from the need of installing any software and will be made publicly available in the near future. At present, it has been designed to work on both conventional and sparse XRF data such as Compressive Sensing, in an embarrassingly parallel manner.

Computational methods are driving high impact microscopy techniques such as ptychography. However, the design and implementation of new algorithms is often a laborious process, as many parts of the code are written in close-to-the-hardware programming constructs to speed up the reconstruction. In this article, we present SciComPty, a new ptychography software framework aiming at simulating ptychography datasets and testing state-of-the-art and new reconstruction algorithms. Despite its simplicity, the software leverages GPU accelerated processing through the PyTorch CUDA interface. This is essential for designing new methods that can readily be employed. As an example, we present an improved position refinement method based on Adam and a new version of the rPIE algorithm, adapted for partial coherence setups. Results are shown on both synthetic and real datasets. The software is released as open-source.

The study of X-Ray fluorescence (XRF) emission spectra is a powerful technique used in applications that range from biology to cultural heritage. Key objectives of this technique include identification and quantification of elemental traces composing the analyzed sample. However, precise derivation of elemental concentration is often hampered by self-absorption of the XRF signal emitted by light constituents. This attenuation depends on the amount of sample present between the radiation source and detection system and allows for the exploitation of self-absorption in order to recover a sample topography. In this work, an x-ray-tracing application based on the use of multiple silicon drift detectors, is introduced to inversely reconstruct a 3D sample with correct topographical landscape, from 2D XRF count rates maps obtained from spectroscopy. The reconstruction was tested on the XRF maps of a simulated sample, which is composed of three cells with different size but similar composition. We propose to use the recovered 3D sample topography in order to numerically compute the self-absorption effects on the X-ray fluorescence radiation, thereby showing that a quantitative correction is possible. Lastly, we present a web application which implements the suggested methodology, in order to demonstrate its feasibility and applicability, available at: https://github.com/ElettraSciComp/xrfstir.

X-ray fluorescence (XRF) is a spectroscopic technique with applications in several fields, such as biology, food science and forensics. Often setups have multi-element detectors in order to improve the signal-to-noise ratio. The relative set of produced spectra have to be aligned to a reference spectrum, in a procedure that is referred to as energy calibration, which is necessary for the fitting. Automated methods fail and a manual procedure is typically employed instead. In this paper, we discuss the signal alignment problem of such systems and we illustrate the preliminary results of a new automated method for linear XRF spectra alignment, which potentially can be used also for other time-series like data.

Soft X-ray microscopy coupled with low energy X-ray fluorescence is a powerful tool for investigating complex biological systems like cells and tissues. Due to certain characteristics of X-ray sources, sample stage motors, and detectors, the examination of large areas at high resolutions is very time consuming, often confining the analysis only to a restricted number of pre-selected representative regions. Here we propose and demonstrate a compressive sensing method that provides an alternative approach for overcoming such limitations and can be applied to different kinds of samples and other microscopy and analytical techniques.

Voice-user interfaces (VUI) use speech recognition tech-nology to enable user interaction with technology. Virtual assistants like Siri, Google Home and Alexa have popular-ized them. The creation of useful VUIs requires the deep understanding of human communication. This manuscript presents a panorama of VUIs and an analysis of the key features needed in order to use them efficiently for Machine and Experiment Control Systems typically used in research infrastructures. Mycroft [1,2] is a leading open source voice assistant that is highly customizable. Tango [3] is an Open Source solution for Supervisory Control and Data Acquisition systems (SCADA) and Distributed Control Systems (DCS). The second part of the manuscript presents and discuss the integration of MyCroft and Tango based control systems for applications in Elettra Sincrotrone Trieste (including experiment control). The potentialities and weakness of the solution, like command uncer-tainty, are presented as challenges. This integration may inspire novel ways of interactions and new developments of interfaces for accelerators, beam lines and large experimental physics control systems.

The high resolution of synchrotron cryo-nano tomography can be easily undermined by setup instabilities and sample stage deficiencies such as runout or backlash. At the cost of limiting the sample visibility, especially in the case of bio-specimens, high contrast nano-beads are often added to the solution to provide a set of landmarks for a manual alignment. However, the spatial distribution of these reference points within the sample is difficult to control, resulting in many datasets without a sufficient amount of such critical features for tracking. Fast automatic methods based on tomography consistency are thus desirable, especially for biological samples, where regular, high contrast features can be scarce. Current off-the-shelf implementations of such classes of algorithms are slow if used on a real-world high-resolution dataset. In this paper, we present a fast implementation of a consistency-based alignment algorithm especially tailored to a multi-GPU system. Our implementation is released as open-source.

Fourier Ptychography is an emerging microscope technique which is increasingly used for biological samples of high importance. The technique allows for stunning resolution without having to cope with a restricted field of view. This result is achieved by combining phase retrieval with an aperture synthesis procedure, indeed casting the technique as one of the most advanced computational imaging investigation methods. Due to the nature of the inverse problem involved (phase retrieval), currently only iterative algorithms can be deployed for reconstruction, thus delaying the image visualization from the acquisition process. In this paper, we propose a deep learning method to seed the iterative reconstruction and obtain a higher quality result in a shorter time. A parameter agnostic CNN is trained to produce an initial estimate for the iterative process. The final reconstructions exhibit a reduced number of artefacts, even for a limited illumination Numerical Aperture. Our method is decisive to relax the design of the illumination array.

Progress in nanotechnology calls for material probing techniques of high sensitivity and resolution. Such techniques are also used for high-impact studies of nanoscale materials in medicine and biology. Soft X-ray microscopy has been successfully used for investigating complex biological processes occurring at micrometric and sub-micrometric length scales and is one of the most powerful tools in medicine and the life sciences. Here, we present the capabilities of the TwinMic soft X-ray microscopy end-station at the Elettra synchrotron in the context of medical and biological imaging, while we also describe novel uses and developments.

X-ray ptychography is an advanced computational microscopy technique, which is delivering exceptionally detailed quantitative imaging of biological and nanotechnology specimens, which can be used for high-precision X-ray measurements. However, coarse parametrisation in propagation distance, position errors and partial coherence frequently threaten the experimental viability. In this work, we formally introduce these actors, solving the whole reconstruction as an optimisation problem. A modern deep learning framework was used to autonomously correct the setup incoherences, thus improving the quality of a ptychography reconstruction. Automatic procedures are indeed crucial to reduce the time for a reliable analysis, which has a significant impact on all the fields that use this kind of microscopy. We implemented our algorithm in our software framework, SciComPty, releasing it as open-source. We tested our system on both synthetic datasets, as well as on real data acquired at the TwinMic beamline of the Elettra synchrotron facility.