This study presents two workflows for leveraging longitudinal computed tomography (CT) datasets when developing deep learning-based detection systems for gradually developing postharvest disorders. Workflow 1 (Longitudinal Benchmarking) benchmarks neural networks by training and testing them on images from different stages of disorder progression. It examines the trade-off between detecting a disorder early or accurately and evaluates whether neural networks can generalize across time points. Workflow 2 (Longitudinal eXplainable Artificial Intelligence (XAI) Heatmaps) provides heatmaps that indicate how changes over time affect the outcomes of neural networks. It uses image registration to align an earlier-acquired image and then uses it as a baseline when calculating the heatmap. The workflows are demonstrated on a dataset of ‘Braeburn’ apples that were CT-scanned multiple times while developing internal browning during controlled-atmosphere (CA) storage and shelf life. The Longitudinal Benchmarking workflow was used to investigate whether images acquired immediately after CA storage can be used to predict the eventual browning after a shelf-life period, which is highly relevant in industrial practice. Moreover, the longitudinal XAI heatmaps avoided artifacts caused by out-of-distribution baselines or identical baseline regions, which occurred with conventional black or zero baselines.

We present a self-supervised deep-learning-based ghost imaging (GI) reconstruction method, which provides a significant improvement in reconstruction quality for noisy acquisitions among unsupervised methods. We present the supporting mathematical framework and results from both theoretical and real data (from a recent synchrotron X-ray GI experiment). Self-supervision removes the need for clean reference data while offering strong noise reduction. This provides the necessary tools for addressing signal-to-noise ratio concerns for GI acquisitions in emerging and cutting-edge low-light GI scenarios. Notable examples include micro- and nano-scale x-ray emission imaging, e.g., x-ray fluorescence imaging of dose-sensitive samples. Their applications include in-vivo and in-operando case studies for biological samples and batteries.

We recently developed a new self-supervised deep-learning-based Ghost Imaging (GI) reconstruction method, which provides unparalleled reconstruction performance for noisy acquisitions among unsupervised methods. Self-supervision removes the need for clean reference data while offering strong noise reduction. This provides the necessary tools for addressing signal-to-noise ratio concerns for GI acquisitions in emerging and cutting-edge low-light GI scenarios. Notable examples include micro- and nano-scale x-ray emission imaging, e.g., x-ray fluorescence (XRF) imaging of dosesensitive samples. Here, we will analyze the performance of our method against the state of the art in unsupervised reconstruction methods for the reconstruction of XRF-GI data.

Deep learning is a potent tool for image segmentation, but it typically demands abundant annotated data for effective training. In scientific domains, such as cell imaging, obtaining annotations for every structure can be prohibitively expensive. Few-shot learning, adapting from one task to another using minimal examples, can alleviate the need for large training data sets. However, there is limited research addressing the particularities of using few-shot learning for cell imaging. Here, we propose a few-shot learning solution designed to be applicable to cell microscopy images. Our method trains feature extractor networks on classes with abundant labelled samples. These feature extractors are then used to generate high-resolution feature maps from the few labelled images of the new class. Finally, we train a perceptron to recombine the feature maps into predicting the new class. On two challenging cell segmentation data sets, we achieve, using five annotated images, Dice scores that are, on average, less than 20

Coherent X-ray microscopy is emerging as a transformative technology for neuronal imaging, with the potential to offer a scalable solution for reconstruction of neural circuits in millimeter sized tissue volumes. Specifically, X-ray holographic nanotomography (XNH) brings together outstanding capabilities in terms of contrast, spatial resolution and data acquisition speed. While recent XNH developments already enabled generating valuable datasets for neurosciences, a major challenge for reconstruction of neural circuits remained overcoming resolving power limits to distinguish smaller neurites and synapses in the reconstructed volumes. Here we present a self-supervised image restoration approach that improves simultaneously spatial resolution, contrast, and data acquisition speed. This enables revealing synapses with XNH, marking a major milestone in the quest for generating connectomes of full mammalian brains. We demonstrate that this method is effective for various types of neuronal tissues and acquisition schemes. We propose a scalable implementation compatible with multi-terabyte image volumes. Altogether, this work brings large scale X-ray nanotomography to a new precision level. ### Competing Interest Statement The authors have declared no competing interest.

X-ray computed tomography (CT) is a powerful imaging tool for defect detection, segmentation and feature extraction in industrial applications as it enables non-destructive evaluation. The presence of artifacts and noise, however, imposes difficulties on the defect detection due to low contrast between defects and material. Different methods, the most recent ones based on deep learning, have been proposed to address both CT artifacts and noise. In this work, three state-of-the-art denoising techniques, namely BM3D, a supervised UNet and Noise2Inverse are compared with a novel approach based on a set of CNNs applied at different stages of the CT data acquisition pipeline. The comparison is done regarding the capabilities of the methods in terms of image quality enhancement as a preprocessing step in CT-based data analysis before performing defect segmentation.

Like in many other research fields, recent developments in computational imaging have focused on developing machine learning (ML) approaches to tackle its main challenges. To improve the performance of computational imaging algorithms, machine learning methods are used for image processing tasks such as noise reduction. Generally, these ML methods heavily rely on the availability of high-quality data on which they are trained. This work explores the application of ML methods, specifically convolutional neural networks (CNNs), in the context of noise reduction for computed tomography (CT) imaging. We utilize a large 2D computed tomography dataset for machine learning to carry out for the first time a comprehensive study on the differences between the observed performances of algorithms trained on simulated noisy data and on real-world experimental noisy data. The study compares the performance of two common CNN architectures, U-Net and MSD-Net, that are trained and evaluated on both simulated and experimental noisy data. The results show that while sinogram denoising performed better with simulated noisy data if evaluated in the sinogram domain, the performance did not carry over to the reconstruction domain where training on experimental noisy data shows a higher performance in denoising experimental noisy data. Training the algorithms with an optimization in the reconstruction domain mapping directly from sinogram to reconstruction significantly improved model performance, emphasizing the importance of matching raw measurement data to high-quality CT reconstructions. The study furthermore suggests the need for more sophisticated noise simulation approaches to bridge the gap between simulated and real-world data in CT image denoising applications and gives insights into the challenges and opportunities in leveraging simulated data for machine learning in computational imaging.

Computed tomography (CT) using synchrotron radiation is a powerful technique that, compared with laboratory CT techniques, boosts high spatial and temporal resolution while also providing access to a range of contrast-formation mechanisms. The acquired projection data are typically processed by a computational pipeline composed of multiple stages. Artifacts introduced during data acquisition can propagate through the pipeline and degrade image quality in the reconstructed images. Recently, deep learning has shown significant promise in enhancing image quality for images representing scientific data. This success has driven increasing adoption of deep learning techniques in CT imaging. Various approaches have been proposed to incorporate deep learning into computational pipelines, but each has limitations in addressing artifacts effectively and efficiently in synchrotron CT, either in properly addressing the specific artifacts or in computational efficiency. Recognizing these challenges, we introduce a novel method that incorporates separate deep learning models at each stage of the tomography pipeline - projection, sinogram and reconstruction - to address specific artifacts locally in a data-driven way. Our approach includes bypass connections that feed both the outputs from previous stages and raw data to subsequent stages, minimizing the risk of error propagation. Extensive evaluations on both simulated and real-world datasets illustrate that our approach effectively reduces artifacts and outperforms comparison methods.

This paper presents a novel method for the reconstruction of high-resolution temporal images in dynamic tomographic imaging, particularly for discrete objects with smooth boundaries that vary over time. Addressing the challenge of limited measurements per time point, we propose a technique that synergistically incorporates spatial and temporal information of the dynamic objects. This is achieved through the application of the level-set method for image segmentation and the representation of motion via a sinusoidal basis. The result is a computationally efficient and easily optimizable variational framework that enables the reconstruction of high-quality 2D or 3D image sequences with a single projection per frame. Compared to current methods, our proposed approach demonstrates superior performance on both synthetic and pseudo-dynamic real X-ray tomography datasets. The implications of this research extend to improved visualization and analysis of dynamic processes in tomographic imaging, finding potential applications in diverse scientific and industrial domains.

Visualizing the internal structure of museum objects is a crucial step in acquiring knowledge about the origin, state, and composition of cultural heritage artifacts. Among the most powerful techniques for exposing the interior of museum objects is computed tomography (CT), a technique that computationally forms a 3D image using hundreds of radiographs acquired in a full circular range. However, the lack of affordable and versatile CT equipment in museums, combined with the challenge of transporting precious collection objects, currently keeps this technique out of reach for most cultural heritage applications. We propose an approach for creating accurate CT reconstructions using only standard 2D radiography equipment already available in most larger museums. Specifically, we demonstrate that a combination of basic X-ray imaging equipment, a tailored marker-based image acquisition protocol, and sophisticated data-processing algorithms, can achieve 3D imaging of collection objects without the need for a costly CT imaging system. We implemented this approach in the British Museum (London), the J. Paul Getty Museum (Los Angeles), and the Rijksmuseum (Amsterdam). Our work paves the way for broad facilitation and adoption of CT technology across museums worldwide. Visualising the structure of museum objects is a crucial step in understanding the origin, state, and composition of cultural heritage artifacts. Here the authors present an approach for creating computed tomography reconstructions using only standard 2D radiography equipment already available in most larger museums.

BACKGROUND: X-ray imaging is widely used for the non-destructive detection of defects in industrial products on a conveyor belt. In-line detection requires highly accurate, robust, and fast algorithms. Deep Convolutional Neural Networks (DCNNs) satisfy these requirements when a large amount of labeled data is available. To overcome the challenge of collecting these data, different methods of X-ray image generation are considered. OBJECTIVE: Depending on the desired degree of similarity to real data, different physical effects should either be simulated or can be ignored. X-ray scattering is known to be computationally expensive to simulate, and this effect can greatly affect the accuracy of a generated X-ray image. We aim to quantitatively evaluate the effect of scattering on defect detection. METHODS: Monte-Carlo simulation is used to generate X-ray scattering distribution. DCNNs are trained on the data with and without scattering and applied to the same test datasets. Probability of Detection (POD) curves are computed to compare their performance, characterized by the size of the smallest detectable defect. RESULTS: We apply the methodology to a model problem of defect detection in cylinders. When trained on data without scattering, DCNNs reliably detect defects larger than 1.3 mm, and using data with scattering improves performance by less than 5%. If the analysis is performed on the cases with large scattering-to-primary ratio (1 < SPR < 5), the difference in performance could reach 15% (approx. 0.4 mm). CONCLUSION: Excluding the scattering signal from the training data has the largest effect on the smallest detectable defects, and the difference decreases for larger defects. The scattering-to-primary ratio has a significant effect on detection performance and the required accuracy of data generation.

Computed tomography (CT) is a widely used non-invasive diagnostic method in various fields, and recent advances in deep learning have led to significant progress in CT image reconstruction. However, the lack of large-scale, open-access datasets has hindered the comparison of different types of learned methods. To address this gap, we use the 2DeteCT dataset, a real-world experimental computed tomography dataset, for benchmarking machine learning based CT image reconstruction algorithms. We categorize these methods into post-processing networks, learned/unrolled iterative methods, learned regularizer methods, and plug-and-play methods, and provide a pipeline for easy implementation and evaluation. Using key performance metrics, including SSIM and PSNR, our benchmarking results showcase the effectiveness of various algorithms on tasks such as full data reconstruction, limited-angle reconstruction, sparse-angle reconstruction, low-dose reconstruction, and beam-hardening corrected reconstruction. With this benchmarking study, we provide an evaluation of a range of algorithms representative for different categories of learned reconstruction methods on a recently published dataset of real-world experimental CT measurements. The reproducible setup of methods and CT image reconstruction tasks in an open-source toolbox enables straightforward addition and comparison of new methods later on. The toolbox also provides the option to load the 2DeteCT dataset differently for extensions to other problems and different CT reconstruction tasks.

Many of the recent successes of deep learning-based approaches have been enabled by a framework of flexible, composable computational blocks with their parameters adjusted through an automatic differentiation mechanism to implement various data processing tasks. In this work, we explore how the same philosophy can be applied to existing “classical” (i.e., non-learning) algorithms, focusing on computed tomography (CT) as application field. We apply four key design principles of this approach for CT workflow design: end-to-end optimization, explicit quality criteria, declarative algorithm construction by building the forward model, and use of existing classical algorithms as computational blocks. Through four case studies, we demonstrate that auto-differentiation is remarkably effective beyond the boundaries of neural-network training, extending to CT workflows containing varied combinations of classical and machine learning algorithms.

A large percentage of apples are affected by internal disorders after long-term storage, which makes them unacceptable in the supply chain. CT imaging is a promising technique for in-line detection of these disorders. Therefore, it is crucial to understand how different disorders affect the image features that can be observed in CT scans. This paper presents a workflow for creating datasets of image pairs of photographs of apple slices and their corresponding CT slices. By having CT and photographic images of the same part of the apple, the complementary information in both images can be used to study the processes underlying internal disorders and how internal disorders can be measured in CT images. The workflow includes data acquisition, image segmentation, image registration, and validation methods. The image registration method aligns all available slices of an apple within a single optimization problem, assuming that the slices are parallel. This method outperformed optimizing the alignment separately for each slice. The workflow was applied to create a dataset of 1347 slice photographs and their corresponding CT slices. The dataset was acquired from 107 'Kanzi' apples that had been stored in controlled atmosphere (CA) storage for 8 months. In this dataset, the distance between annotations in the slice photograph and the matching CT slice was, on average, 1.47 ± 0.40 mm. Our workflow allows collecting large datasets of accurately aligned photo-CT image pairs, which can help distinguish internal disorders with a similar appearance on CT. With slight modifications, a similar workflow can be applied to other fruits or MRI instead of CT scans.

X-ray imaging can be efficiently used for high-throughput in-line inspection of industrial products. However, designing a system that satisfies industrial requirements and achieves high accuracy is a challenging problem. The effect of many system settings is application-specific and difficult to predict in advance. Consequently, the system is often configured using empirical rules and visual observations. The performance of the resulting system is characterized by extensive experimental testing. We propose to use computational methods to substitute real measurements with generated images corresponding to the same experimental settings. With this approach, it is possible to observe the influence of experimental settings on a large amount of data and to make a prediction of the system performance faster than with conventional methods. We argue that a high accuracy of the image generator may be unnecessary for an accurate performance prediction. We propose a quantitative methodology to characterize the quality of the generation model using Probability of Detection curves. The proposed approach can be adapted to various applications and we demonstrate it on the poultry inspection problem. We show how a calibrated image generation model can be used to quantitatively evaluate the effect of the X-ray exposure time on the performance of the inspection system.

3D imaging methods are increasingly employed in cultural heritage research to analyse and document objects in museum collections. In this work, we provide an interactive visualisation plugin for the open-source software Blender, to combine and inspect two complementary 3D imaging modalities: CT images, which capture the interior; and surface scans, which capture the exterior. 3D CT scan data can be visualised, both as volumetric representation and as orthogonal slices, and a 3D surface scan can be registered onto the CT data. It allows users to simultaneously and interactively inspect these modalities and to virtually cut through an object. It also provides tools for generating output images and videos for research and public outreach purposes. The plugin workflow was applied to four case studies from the collections of the Rijksmuseum, Amsterdam, and the British Museum, London. The plugin is published open-source together with detailed guidelines and a practice dataset.

Computed tomography (CT) is a powerful non-invasive tool to analyze cultural heritage objects by allowing museum professionals to obtain 3D information about the objects’ interior. These insights can help with the conservation or restoration of the objects, as well as provide contextual information on the objects’ history or making process. Cultural heritage objects exist in a wide variety and have characteristics which present challenges for CT scanning: multi-scale internal features, a diversity of sizes and shapes, and multi-material objects. Because X-ray absorption is related to the density, thickness of the material, and atomic composition, the challenges are greater when the object consists of multiple different materials with varying densities. This is especially true for cases with extreme density contrasts such as that between metals and textiles. An untailored acquisition of CT scans of multi-material objects can lead to reduced image quality and heavy visual errors called image artifacts, which can influence the perception or representation of information. A tailored acquisition can reduce these artifacts and lead to a higher information gain. In this work, we firstly discuss how the X-ray beam properties and the beam-object interaction influence CT image formation and how to use filters to manipulate the emitted X-ray beam to improve image quality for multi-material objects. We showcase that this can be achieved with limited resources in a low-cost DIY fashion with thin sheets of metal as filters, 3D-printed filter frames and a filter holder. Secondly, we give a qualitative analysis of the influence of the CT acquisition parameters illustrated with two case study objects from the textile collection of the Rijksmuseum, Amsterdam, The Netherlands. With this we provide insights and intuitions on tailoring the CT scan to the cultural heritage objects. Thirdly, we extract a general concept of steps for museum professionals to design an object-tailored CT scan for individual cases.

Computed tomography (CT) is a widely used non-invasive medical imaging technique for disease diagnosis. The diagnostic accuracy is often affected by image resolution, which can be insufficient in practice. For medical CT images, the through-plane resolution is often worse than the in-plane resolution and there can be overlap between slices, causing difficulties in diagnoses. Self-supervised methods for through-plane resolution enhancement, which train on in-plane images and infer on through-plane images, have shown promise for both CT and MRI imaging. However, existing self-supervised methods either neglect overlap or can only handle specific cases with fixed combinations of resolution and overlap. To address these limitations, we propose a self-supervised method called SR4ZCT. It employs the same off-axis training approach while being capable of handling arbitrary combinations of resolution and overlap. Our method explicitly models the relationship between resolutions and voxel spacings of different planes to accurately simulate training images that match the original through-plane images. We highlight the significance of accurate modeling in self-supervised off-axis training and demonstrate the effectiveness of SR4ZCT using a real-world dataset.

Deep-learning algorithms for cell segmentation typically require large data sets with high-quality annotations to be trained with. However, the annotation cost for obtaining such sets may prove to be prohibitively expensive. Our work aims to reduce the time necessary to create high-quality annotations of cell images by using a relatively small well-annotated data set for training a convolutional neural network to upgrade lower-quality annotations, produced at lower annotation costs. We investigate the performance of our solution when upgrading the annotation quality for labels affected by three types of annotation error: omission, inclusion, and bias. We observe that our method can upgrade annotations affected by high error levels from 0.3 to 0.9 Dice similarity with the ground-truth annotations. We also show that a relatively small well-annotated set enlarged with samples with upgraded annotations can be used to train better-performing cell segmentation networks compared to training only on the well-annotated set. Moreover, we present a use case where our solution can be successfully employed to increase the quality of the predictions of a segmentation network trained on just 10 annotated samples.