Microscopy (Oxford, England)最新文献

Traditional three-dimensional (3D) reconstruction is labor-intensive owing to manual segmentation; this can be addressed by developing artificial intelligence (AI)-driven automated segmentation. However, it is limited by a lack of user-friendly tools for morphologists. We present a workflow for 3D reconstruction using our AI-powered segmentation tool. Specifically, we developed an interactive toolset, 'Seg & Ref', to overcome the abovementioned challenges by enabling AI-powered segmentation and easy mask editing without requiring a command-line setup. We demonstrated a 3D reconstruction workflow using serial sections of a Carnegie Stage 15 human embryo. Automated segmentation (Step 1) was performed using the graphical user interface, 'SAM2 GUI for Img Seq', which utilizes the Segment Anything Model 2 and supports interactive segmentation through a web-based interface. Users specify target structures via box prompts, and the results are propagated across all images for batch segmentation. The segmentation masks were reviewed and corrected (Step 2) using 'Segment Editor PP', a PowerPoint-based tool enabling interactive mask refinement. Finally, the corrected masks were imported into the 3D Slicer application for reconstruction (Step 3). Our 3D reconstruction visualized key structures, including the spinal cord, veins, aorta, mesonephros, gut, heart, trachea, liver and peritoneal cavity. The reconstructed models accurately represented their spatial relationships and morphologies. This provides a labor-saving approach for 3D reconstruction workflows owing to their optimization for serial sections, versatility and accessibility without programming expertise. Therefore, morphological research can be enhanced by precise segmentation using intuitive and user-friendly interfaces of 'Seg & Ref'.

The mechanism of voltage contrast formation under ultra-low landing energy condition is discussed, by which a binder contained in lithium-ion battery anode material has been visualized with high contrast. Since the anode material is a complex experimental system with multiple contrast formation factors, a standard sample simulating it was fabricated for simplification. The binder was observed to be darker than the substrate at landing energies of 30-50 eV. The binder exhibited a distinct appearance reflecting its shape (in the 3D-particle mode) at 20 eV. The mirroring phenomenon occurred at 10 eV, in which the primary electrons bounced off the sample before irradiating the surface. The surface potential at the electron beam irradiation moment was presumed to affect the contrast formation, but direct measurement of it was difficult. Thus, the sample was transferred to an Atomic Force Microscope without exposure to the atmosphere to measure the 'residual' potential of the binder in KPFM mode after the SEM observations. Under darker binder observed conditions of 30-50 eV, KPFM measured residual potential was positive relative to the substrate. Under conditions of the 3D-particle mode at 20  eV and the mirroring phenomenon at 10 eV, the residual potentials were negative. Therefore, a correlation between the behavior of the voltage contrast and the residual potential was obtained. Finer landing-energy step measurement revealed hysteresis responses of voltage contrast and the residual potential to the landing energy. The Cause of the hysteresis was discussed.

Scientific research relies on microscopy. However, manual image acquisition and analysis are inefficient and susceptible to errors. Fully automated workflows are often task-specific, and current AI-based systems are costly and may face difficulties in new scenarios. Here, we introduce a semi-automated system utilizing macro keyboards to streamline workflows. Programming multi-action keys for tasks such as focusing, image capture and data analysis reduces the manual input, boosting efficiency and accuracy. This intuitive system saves time for both experienced users and trainees. This cost-effective solution improves accessibility, flexibility and usability, supporting not only diverse imaging applications but also broader scientific instrumentation processes.

Imaging performance with 120 and 200 keV electrons was evaluated with an integration-type silicon-on-insulator pixel detector called INTPIX4 installed in a conventional transmission electron microscope. We demonstrated that single-electron events can be detected with INTPIX4 quantitatively. The gain and signals of single-electron events were measured. On the basis of the results, the yields of the collected charge for 120 and 200 keV electrons were estimated to be 96±5% and 97±5%, respectively. The modulation transfer function and detective quantum efficiency were also measured. INTPIX4 was clarified to have high detection efficiency and high sensitivity. We also found that it is necessary to use electron beams with energies less than 120 keV for INTPIX4 because multiple scattering of primary electrons at the silicon sensor degrades image resolution. This detector is expected to be applicable to low-dose observations in transmission electron microscopy.

Improvement of a commercially available soft X-ray emission spectrometer was tested by introducing a fine-pixel-sized CMOS detector. The peak width of Mg Kα-emission was reduced to one-fourth of that obtained by the CCD detector presently used. Furthermore, the differences in the energy positions of satellite lines of Mg Kα- and also Kβ-emission profiles of Mg and MgO were observed. O K-emission profile of MgO exhibited a few structures reflecting the chemical bonding state. This spectrometer easily discriminated the intensity profiles of Fe Lα,β-emission reflecting the chemical bonding states of Fe atoms in Fe, FeO, Fe3O4 and Fe2O3.

According to theoretical predictions, local strain in the bent regions of flexible nanowires can alter their electronic structure. However, the experimental validation of such strain-induced effects remains elusive. In this study, we established a clear correlation between local structural deformation and electronic properties in bent hexagonal-WO3 nanowires using four-dimensional scanning transmission electron microscopy and electron energy loss spectroscopy. Although a simple geometric bending model predicts an expansion of the (0001) lattice spacing on the outer side of the bend, our direct observations revealed a larger expansion than predicted. This lattice expansion was accompanied by a significant reduction in bandgap energy. We employed density functional theory calculations and crystal orbital Hamilton population analyses to provide a theoretical framework for these findings. These results provide direct experimental evidence of strain-induced modulation of the electronic structure in metal oxide nanowires.

Electron spectroscopy implemented in electron microscopes provides high spatial resolution, down to the atomic scale, of the chemical, electronic, vibrational and optical properties of materials. In this review, we will describe how temporal coincidence experiments in the nanosecond to femtosecond range between different electron spectroscopies involving photons, inelastic electrons and secondary electrons can provide information bits not accessible to independent spectroscopies. In particular, we will focus on nano-optics applications. The instrumental modifications necessary for these experiments are discussed, as well as the perspectives for these coincidence techniques.

Knife-mark noise often arises in microscopy of materials. Leveraging their simple textures relative to natural images, we simulate knife-marked micrographs and train a deep network without labeled real data. The resulting model surpasses conventional methods, removing artifacts while preserving structure, demonstrating simulation-driven learning as a practical materials-science solution in research. Accurate quantitative analysis of material microstructures from images is often hindered by noise and artifacts generated during sample preparation. While deep learning is a promising approach for this challenge, preparing the large amount of "supervised data" (labeled real images) required for training poses a significant barrier in material science. This study proposes and validates a simulation-driven learning paradigm where a deep learning model is trained exclusively on simulated images that mimic the key features of target structures and noise, serving as a powerful solution to this data scarcity problem. As a specific case study, we applied this paradigm to the removal of "knife-mark noise" from cross-sectional images of rubber materials to enable accurate filler region segmentation. In evaluations using simulated data, the proposed method showed superior performance across all the metrics (PSNR, SSIM, and MAE) compared with conventional methods such as the median filter and TV reconstruction, as well as a U-Net model trained on general-purpose Gaussian noise. More importantly, the model also performed effectively on real images, despite being trained solely on simulated data. It successfully removed both knife-marks and material-derived background textures, which demonstrates the viability of simulation-driven learning to overcome the need for manually annotated datasets. This work highlights the power of task-specific simulations as a practical alternative to manual data annotation in quantitative materials analysis.

Dynamics in liquids and glasses can be assessed using X-ray photon correlation spectroscopy or electron correlation microscopy, which involves measuring the temporal changes in diffraction patterns. Two methods are commonly used to evaluate these temporal changes: one-time correlation function or two-time correlation function. However, the specific characteristics of these methods have not been thoroughly studied. In this study, we investigated the differences between these methods and found that the two-time correlation function can measure dynamics for longer periods than the method relying on the one-time correlation function. Additionally, we demonstrated that the two-time correlation function exhibits a weak dependence on the amount of dose applied.

Atomic force microscopy (AFM) has developed remarkably in recent years, and its measurement environment has been extended not only to ultrahigh vacuum and air, but also to liquids. Since the solid-liquid interface is the site of various reactions, such as crystal growth and catalytic reactions, its atomic-scale analysis is crucially important. Although AFM analyses in various liquids, such as aqueous solutions, organic solvents, and ionic liquids, have been reported, there have been no studies of AFM analysis in molten metals. One of the reasons for this is the opacity of molten metals. Achieving AFM analysis in molten metal is expected to provide new insights into metallurgy. In this review, AFM that can analyze in molten metal is presented. The key innovation is the utilization of an AFM sensor employing a quartz tuning fork, the so-called qPlus sensor, instead of a silicon cantilever. In addition to the technical fundamentals of AFM in molten metal, we present two applications: in-situ and atomic-resolution analysis of alloy crystal growth processes and measurements of two-body interaction forces.