EELS elemental mapping with unconventional methods I. Theoretical basis: Image analysis with multivariate statistics and entropy concepts

EELS elemental mapping with unconventional methods II. Applications to biological specimens

We investigated the estimation of the ensemble channelized Hotelling observer (CHO) signal-to-noise ratio (SNR) for ordered-subset (OS) image reconstruction using noisy projection data. Previously, we computed the ensemble CHO SNR using a method for approximating the channelized covariance of OS reconstruction, which requires knowledge of the noise-free projection data. Here, we use a “plug-in” approach, in which noisy data is used in place of the noise-free data in the aforementioned channelized covariance approximation. Additionally, we evaluated the use of smoothing of the noisy projections before use in the covariance approximation. Additionally, we evaluated the use of smoothing of the noisy projections before use in the covariance calculation. The task was detection of a 10% contrast Gaussian signal within a slice of the MCAT phantom. Simulated projections of the MCAT phantom were scaled and Poisson noise was added to create 100 noisy signal-absent data sets. Simulated projections of the scaled signal were then added to the noisy background projections to create 100 noisy signal-present data set. These noisy data sets were then used to generate 100 estimates of the ensemble CHO SNR for reconstructions at various iterates. For comparison purposes, the same calculation was repeated with the noise-free data. The results, reported as plots of the average CHO SNR generated in this fashion, along with 95% confidence intervals, demonstrate that this approach works very well, and would allow optimization of imaging systems and reconstruction methods using a more accurate object model (i.e., real patient data).

Seismic data denoising is a critical component of seismic data processing, yet effectively removing erratic noise, characterized by its non-Gaussian distribution and high amplitude, remains a substantial challenge for conventional methods and deep-learning (DL) algorithms. Supervised learning frameworks typically outperform others, but they require pairs of noisy data sets alongside corresponding clean ground truth, which is impractical for real-world seismic data sets. In contrast, unsupervised learning (UL) methods, which do not rely on ground truth during training, often fall short in performance when compared with their supervised or traditional denoising counterparts. Moreover, current unsupervised DL methods fail to address the specific challenges posed by erratic seismic noise adequately. This paper introduces a novel zero-shot unsupervised DL framework designed specifically to mitigate random and erratic noise, with a particular emphasis on blended noise. Drawing inspiration from Noise2Noise (N2N) and data augmentation principles, we develop a robust self-supervised denoising network called robust Noiser2Noiser. Our approach eliminates the need for paired noisy and clean data sets as required by supervised methods or paired noisy data sets as in N2N. Instead, our framework relies solely on the original noisy seismic data set. Our methodology generates two independent recorrupted data sets from the original noisy data set, using one as the input and the other as the training target. Subsequently, we use a DL-based denoiser, denoising convolutional neural network, for training purposes. To address various types of random and erratic noise, the original noisy data set is recorrupted with the same noise type. Detailed explanations for generating training input and target data for blended data are provided. We apply our network to synthetic and real marine data examples, demonstrating significantly improved noise attenuation performance compared with traditional denoising methods and state-of-the-art UL Codes are available on https://github.com/Ji-seismic/N2N_deblending.

The storage structures of spores of Gigaspora margarita Becker & Hall, an arbuscular mycorrhizal fungus, were analyzed by a transmission electronic microscope (TEM) and electron energy loss spectroscopy (EELS). The spores were hydrated, cryofixed using high pressure freezing, submitted to freeze substitution, embedded and prepared for TEM observations. The cryotechniques revealed that the storage structures in the spores are composed of lipids, electron dense granules (EDG), protein bodies and glycogen particles. The EELS detected the presence of nitrogen (N) and carbon (C) in these storage structures and in Bacteria-like organisms (BLO). Some amount of P was detected in the vacuoles and in BLO. The TEM and EELS techniques indicate the presence of storage structures in the fungal spore, and some essential elements (N, P and C) in these structures.

Electron energy loss spectroscopy (EELS) is an analytical technique that can provide the structural, physical and chemical information of materials. The EELS spectra can be obtained by combining with TEM at sub-nanometer spatial resolution. However, EELS spectral information can’t be obtained easily because in order to interpret EELS spectra, we need to refer to and/or compare many reference data with each other. And in addition to that, we should consider the different experimental variables used to produce each data. Therefore, reliable and easily interpretable EELS standard reference data are needed.Our Electron Energy Loss Data Center (EELDC) has been designated as National Standard Electron Energy Loss Data Center No. 34 to develop EELS standard reference (SR) data and to play a role in dissemination and diffusion of the SR data to users. EELDC has developed and collected EEL SR data for the materials required by major industries and has a total of 82 EEL SR data. Also, we have created an online platform that provides a one-stop-place to help users interpret quickly EELS spectra and get various spectral information. In this paper, we introduce EEL SR data, the homepage of EELDC and how to use them.

We investigated the local structure at the interface between titanium oxide (TiO x ) and crystalline Si (c-Si) with silicon oxide (SiOx) interlayers by employing the electron energy loss (EEL) spectroscopy to clarify the origin of the high performance TiO x passivating layer deposited by atomic layer deposition. Five kinds of TiO x /SiO x /c-Si structures were fabricated and the best passivation performance was achieved by the SiO x interlayer formed by nitric acid at room temperature. EEL spectra revealed that Ti-contained more stoichiometric SiO x including few oxygen vacancies and Ti-O-Si bonding would be formed after forming gas annealing, resulting in the enhancement of passivation performance.

Li-ion batteries have yet to reach their full application potential due to their inherent cycling performance issues, such as voltage instability and capacity fading, while fundamental questions remain unanswered regarding the mechanisms responsible. However, Mn-based layered oxides continue to be promising candidates for cathodes in high-energy-density Li batteries given their high voltage and high discharge capacities. A significant impediment to gaining insight into these layered oxides lies in their complexity and inhomogeneity. Thus, the focus of the present contribution will be on the parent material, Li2MnO3, which shows cycling characteristics similar to its more complex offspring. Scanning transmission electron microscopy (STEM) and spectroscopy are used to characterize structural and electronic properties of both pristine and cycled material. STEM-based methods are quickly becoming the most promising characterization tools for these and similar materials, owed largely to the wide-range of techniques available on advanced STEM instruments, including the direct imaging of both heavy and light elements, and both energy-dispersive X-ray (EDX) and electron energy loss (EEL) spectroscopies. Imaging modes such as high/low angle annular dark field (H/LAADF) and annular bright field (ABF) are exploited to image heavy atomic columns, strain contrast, and light atomic columns, respectively. Additionally, electron energy loss spectroscopy along with calculations based on density functional theory are used to probe the local electronic structure by monitoring the O K- and Mn L-edges, which can be used to track changes to both the O content and the Mn valence. Thus, the focus will remain on the structural and electronic evolution of the pristine layered oxide explored by combining spectroscopy and atomic-scale imaging with various in situ microscopy techniques and ex situ electrochemical cycling. Specifically, features such as the atomic ordering of Mn/Li atoms, O vacancy evolution, and Mn valence will be of particular interest.

The aim of this article is to present a dynamic record linkage solution using high level machine learning to solve the problems associated with noisy, inconsistent and dynamic data sets. Classic deterministic and probabilistic models fail to work in such environments because they are static and based on assumptions. Incorporating supervised learning algorithm, active learning, and ensemble techniques such as random forests and boosting allows the proposed structure to change in response to different data types and thus become more accurate and scaleable. Some of its highlights are feature selection to ensure match accuracy, clustering to support noisy inputs, and active learning to minimize reliance on large labeled datasets. Simulations with real data, such as government or healthcare data, show that compared with the traditional approach, linkage is significantly more accurate and efficient. The flexible model led to 15% higher F1-scores on noisy datasets and was scalable across large data sets. This work demonstrates how adaptive machine learning can revolutionize the modern record linkage tasks and provides a powerful and effective solution to ever-changing data conditions.

Introduction Transmission electron energy loss spectroscopy (EELS) consists of measuring energy loss dispersion of inelastically scattered high-energy electrons transmitted through a thin film. The high-energy electrons, which interact with the electrons in the solid, lose a certain amount of energy and transfer momentum to the solid. Because of the energy and momentum conservation rules, energy loss and the corresponding momentum-transfer ( q ) of the probed electron represents the energy and the momentum of the electronic excitations in solids. Although some optical techniques, such as soft X-ray absorption and optical reflectance measurements, provide comparative information about solids with higher energy resolution, EELS enjoys several unique advantages over optical spectroscopies. First of all, unlike optical reflectance measurements which are sensitive to the surface condition of the sample, the transmitted EELS represents the bulk properties of the material. Secondly, EELS spectra can be measured with q along specific controllable directions and thus, can be used to study the dispersion of plasmons, excitons, and other excitations [8.1–8.5]. Such experiments offer both dynamics as well as symmetry information about the electronic excitations in solids. In addition, the capability to probe the electronic structure at finite momentum-transfer also allows one to investigate the excited monopole or quadrupole transitions, which cannot be directly observed by conventional optical techniques limited by the dipole selection rule. Because of the significant energy spread of conventional TEM electron sources (e.g. LaB 6 , W-hairpin filaments with Δ E ∼ 1–2 eV), EELS measurements to investigate the electronic structure of solids have been generally limited to dedicated electron energy loss spectrometers with energy resolutions ∼0.1 eV.

Accurate blood glucose level prediction is crucial for effective diabetes management, enabling timely interventions and reducing complications. The limitations of traditional machine learning models like Random Forest and XGBOOST include the inability to deal with noisy data and datasets with imbalances. This study proposes a hybrid ensemble model that combines Random Forest for robust feature selection and initial predictions with XGBOOST to improve accuracy by focusing on instances that have been misclassified. The hybrid approach improves the handling of imbalanced and noisy datasets, achieving better accuracy and generalization. The model's superiority over standalone machine learning models is demonstrated by experimental results on publicly accessible datasets, highlighting its potential as a reliable method for predicting blood glucose levels and supporting diabetes management.

Electron energy-loss spectroscopy and electron probe X-ray microanalysis of exhaust aerosols in slurry nebulization inductively coupled plasma atomic spectrometry for ceramic powders

The effect of a 10 nm Ce coating layer on long term oxidation behaviour (up to 3000 hours) of a Fe–22%Cr ferritic stainless steel for solid oxide fuel cell (SOFC) interconnect application is investigated. Scanning transmission electron microscopy (STEM) and electron energy loss spectroscopy (EELS) analysis showed segregation of Ce to the grain boundaries of Cr,Mn spinel layer adjacent to the scale-gas interface. The findings in this study are in line with the grain boundary blocking model for the Ce effect mechanism. However, segregation of Ce was observed at the grain boundaries of the (Cr,Mn) spinel in the vicinity of the scale-gas interface. No evidence of Ce segregation was found in the grain boundaries of the chromia layer, neither any Ce rich particle was observed in the chromia layer after longer exposure times.

We have investigated the chemical mechanism of surface enhanced Raman scattering (SERS) on an atomically smooth metal surface using electron energy loss spectroscopy (EELS) and molecular spectroscopy simulations. The EEL spectra of pyromellitic dianhydride (PMDA) adsorbed on Cu(100) and Cu(111) are reported. Simulations of the surface-enhanced Raman spectra and electron energy loss spectra (EELS) of pyromellitic dianhydride adsorbed on Cu(100) and Cu(111) are reported. The surface enhanced Raman spectra [J. Chem. Soc. Faraday Trans. 92, 4775 (1996)] and the EEL spectra are shown to be sensitive to crystal face. The relevant excited state observed in the EEL spectrum is not intrinsic to molecular PMDA, but results from chemisorption. The Raman spectra are sensitive to the incident laser polarization on both the (100) and (111) surfaces but in different ways. These observations are shown to be a result of the excited state potential energy surface having different shape, and the respective transition dipole moment having a different orientation on the two crystal faces. The nuclear coordinate dependence of the electronic transition dipole moment produces mode selective, polarization dependent Raman scattering cross sections. Based upon this observation we conclude that the transition dipole moment function that couples the ground electronic state to the resonant excited electronic state is also sensitive to the structure of the metal surface.

Microwave Fixation and Localization of Calcium in Synaptic Terminals Using X-Ray Microanalysis and Electron Energy Loss Spectroscopy Imaging

Conventional parallel detectors for Electron Energy Loss Spectroscopy (EELS) have been mainly based upon systems using linear photodiode arrays in a conjugate image plane of an electron spectrometer. We have developed a unique two dimensional charge coupled device (CCD) camera system which can be used as a detector for EEL spectroscopy and imaging, utilizing high sensitivity, high resolution CCD's, which are typically used in medial or astronomic imaging.The present detector system is based upon a Tektronics TK512M 512 x 512 pixel CCD array, (figure 1) which is optically coupled to a YAG:Ce single crystal scintillator. This CCD imaging system views an electron energy loss spectrum which is magnified by a quadrupole doublet lens attached to a Gatan 607 electron spectrometer on a Philips EM420 TEM as is illustrated in figure 2. The CCD controller, detector head electronics and electron optics were developed at Argonne specifically for high speed data acquisition and allow the recording of complete spectra in as short a time as 10 μsec or approximately 103times faster than the typical 1024 pixel photodiode arrays’ thus allowing the potential for time resolved spectroscopy.