Sparsity-Based Super Resolution for SEM Images

Rapid identification of the aging time of Liupao tea using AI-multimodal fusion sensing technology combined with analysis of tea polysaccharide conjugates

Identifying the aging time of Liupao Tea (LPT) presents a persistent challenge. We utilized an AI-Multimodal fusion method combining FTIR, E-nose, and E-tongue to discern LPT's aging years. Compared to single-source and two-source fusion methods, the three-source fusion significantly enhanced identifying accuracy across all four machine learning algorithms (Decision tree, Random forest, K-nearest neighbor, and Partial least squares Discriminant Analysis), achieving optimal accuracy of 98-100 %. Physicochemical analysis revealed monotonic variations in tea polysaccharide (TPS) conjugates with aging, observed through SEM imaging as a transition from lamellar to granular TPS conjugate structures. These quality changes were reflected in FTIR spectral characteristics. Two-dimensional correlation spectroscopy (2D-COS) identified sensitive wavelength regions of FTIR from LPT and TPS conjugates, indicating a high similarity in spectral changes between TPS conjugates and LPT with aging years, highlighting the significant role of TPS conjugates variation in LPT quality. Additionally, we established an index for evaluating quality of aging, which is sum of three fingerprint peaks (1029 cm-1, 1635 cm-1, 2920 cm-1) intensities. The index could effectively signify the changes in aging years on macro-scale (R2 = 0.94) and micro-scale (R2 = 0.88). These findings demonstrate FTIR's effectiveness in identifying aging time, providing robust evidence for quality assessment.

PCL/Yam Polysaccharide nanofibrous membranes loaded with self-assembled HP-β-CD/ECG inclusion complexes for food packaging

In this study, Polycaprolactone (PCL)/Yam Polysaccharide (YP) fiber membranes loaded the ultrasound-mediated assembly of 2-Hydroxypropyl-β-cyclodextrin (HP-β-CD)/Epicatechin gallate (ECG) inclusion complexes were prepared by electrospinning technology for food packaging. Morphology, infrared spectroscopy and X-ray diffraction results showed that the inclusion complexes were successfully assembled. With the addition of inclusion complexes, the average diameter of the fibers increased from 2480.96 to 10179.12 nm, the crystallinity decreased, the thermal stability improved, the hydrophilicity enhanced, and the water vapor permeability enhanced. Meanwhile, thermogravimetry and differential scanning calorimetry results showed that the inclusion complexes formed hydrogen bonds between the fibers, which improved the thermal stability, but the mechanical behavior suffered a certain loss. In addition, the fiber membrane could continuously release ECG within 240 h, which showed excellent antibacterial effects both in vitro and in vivo. These results indicated that the fiber film developed based on electrospinning had a broad application prospect in food packaging.

Reduction of sulfur in fuel oil using Fe2O3 hybrid nanoadsorbent by solvent deasphalting and optimization of operational parameters with CCD

The present study investigated and tested the effect of adding three types of nanoadsorbents (multi-walled carbon nanotubes (MWCNT)) in pure form, multi-walled carbon nanotubes with Fe2O3 particles (MWCNT-Fe2O3) hybrid, and Silanated-Fe2O3 hybrid to heavy fuel oil to reduce sulfur using a deasphalting process with solvent. First, all three types of nanoadsorbents were synthesized. Then, the Central Composite Design (CCD) method was used to identify the parameters effective in deasphalting, such as the type of nanoadsorbent, the weight percentage of nanoadsorbent, and the solvent-to-fuel ratio, and to obtain their optimal values. Based on the optimization result, under laboratory temperature and pressure conditions, the highest percentage of sulfur reduction in deasphalted fuel (DAO) was obtained by adding 2.5% by weight of silanated-Fe2O3 nano-adsorbent and with a solvent-to-fuel ratio of 7.7 (The weight percentage of sulfur in DAO decreased from 3.5% by weight to 2.46%, indicating a decrease of 30%). Additionally, by increasing the temperature to 70 °C, in optimal conditions, the results revealed that the remaining sulfur percentage in DAO decreased to 2.13% by weight, indicating a decrease of 40%. Synthesized nanoadsorbents and asphaltene particles adsorbed on the surfaces of nanoadsorbents were evaluated by XRD, FTIR, FESEM, and TEM techniques.

Enhancing performance of electron holography with mathematical and machine learning-based denoising techniques

Electron holography is a useful tool for analyzing functional properties, such as electromagnetic fields and strains of materials and devices. The performance of electron holography is limited by the 'shot noise' inherent in electron micrographs (holograms), which are composed of a finite number of electrons. A promising approach for addressing this issue is to use mathematical and machine learning-based image-processing techniques for hologram denoising. With the advancement of information science, denoising methods have become capable of extracting signals that are completely buried in noise, and they are being applied to electron microscopy, including electron holography. However, these advanced denoising methods are complex and have many parameters to be tuned; therefore, it is necessary to understand their principles in depth and use them carefully. Herein, we present an overview of the principles and usage of sparse coding, the wavelet hidden Markov model and tensor decomposition, which have been applied to electron holography. We also present evaluation results for the denoising performance of these methods obtained through their application to simulated and experimentally recorded holograms. Our analysis, review and comparison of the methods clarify the impact of denoising on electron holography research.

Oriented artificial niche provides physical-biochemical stimulations for rapid nerve regeneration

Skin wound is always accompanied with nerve damage, leading to significant sensory function loss. Currently, the functional matrix material based stem cell transplantation and in situ nerve regeneration are thought to be effective strategies, of which, how to recruit stem cells, retard senescence, and promote neural differentiation has been obstacle to be overcome. However, the therapeutic efficiency of the reported systems has yet to be improved and side effect reduced. Herein, a conduit matrix with three-dimensional ordered porous structures, regular porosity, appropriate mechanical strength, and conductive features was prepared by orienting the freezing technique, which was further filled with neural-directing exosomes to form a neural-stimulating matrix for providing hybrid physical-biochemical stimulations. This neural-stimulating matrix was then compacted with methacrylate gelatin (GelMA) hydrogel thin coat that loaded with chemokines and anti-senescence drugs, forming a multi-functional artificial niche (termed as GCr-CSL) that promotes MSCs recruitment, anti-senescence, and neural differentiation. GCr-CSL was shown to rapidly enhances in situ nerve regeneration in skin wound therapy, and with great potential in promoting sensory function recovery. This study demonstrates proof-of-concept in building a biomimetic niche to organize endogenous MSCs recruitment, differentiation, and functionalization for fast neurological and sensory recovery.

Enhanced Scanning Electron Microscopy Using Auto-Optimized Image Restoration With Constrained Least Squares Filter for Nanoscience

The growing demands of nanoscience require the continuous improvement of visualization methods. The imaging performance of scanning electron microscopy (SEM) is fundamentally limited by the point spread function of the electron beam and degrades because of noise. This paper proposes an auto-optimization algorithm based on deconvolution for the restoration of SEM images. This algorithm uses a constrained least squares filter and does not dependent on the user's experience or the availability of nondegraded images. The proposed algorithm improved the quality of the SEM images of 10-nm Au nanoparticles, and achieved balance among the sharpness, contrast-to-noise ratio (CNR), and image artifacts. For the SEM image of 100-nm pitched line patterns, the analysis of the spatial frequencies allowed the 2.5-fold improvement of the intensity of 4-nm information, and the noise floor decreased approximately 32 times. Along with the results obtained by the application of the proposed algorithm to images of tungsten disulfide (WS2) flakes, carbon nanotubes (CNTs), and HeLa cells, the evaluation results confirm that the proposed algorithm can enhance the SEM imaging of nanoscale features that lie close to the microscope's resolution limit.

An Adversarial Learning Approach for Super-Resolution Enhancement Based on AgCl@Ag Nanoparticles in Scanning Electron Microscopy Images

Scanning electron microscopy (SEM) plays a crucial role in the characterization of nanoparticles. Unfortunately, due to the limited resolution, existing imaging techniques are insufficient to display all detailed characteristics at the nanoscale. Hardware-oriented techniques are troubled with costs and material properties. Computational approaches often prefer blurry results or produce a less meaningful high-frequency noise. Therefore, we present a staged loss-driven neural networks model architecture to transform low-resolution SEM images into super-resolved ones. Our approach consists of two stages: first, residual channel attention network (RCAN) with mean absolute error (MAE) loss was used to get a better peak signal-to-noise ratio (PSNR). Then, discriminators with adversarial losses were activated to reconstruct high-frequency texture features. The quantitative and qualitative evaluation results indicate that compared with other advanced approaches, our model achieves satisfactory results. The experiment in AgCl@Ag for photocatalytic degradation confirms that our proposed method can bring realistic high-frequency structural detailed information rather than meaningless noise. With this approach, high-resolution SEM images can be acquired immediately without sample damage. Moreover, it provides an enhanced characterization method for further directing the preparation of nanoparticles.

Computational Evaluation of Sparse Coding on off-axis Electron Holograms: Comparison Between Charge-Coupled Device and Direct-Detection Device Cameras

The effectiveness of sparse coding for image inpainting and denoising of off-axis electron holograms was examined computationally based on hologram simulations according to considerations of two types of electron detectors, namely, charge-coupled device (CCD) and direct-detection device (DDD) cameras. In this simulation, we used a simple-phase object with a phase step such as a semiconductor p-n junction and assumed that the holograms recorded by the CCD camera include shot noise, dark-current, and read-out noise, while those recorded by the DDD camera include only shot noise. Simulated holograms with various electron doses were sparsely coded. Even though interference fringes cannot be recognized in the simulated CCD and DDD holograms when subjected to electron doses (per pixel) equal to 1 and 0.01, respectively, both the corresponding sparse-coded holograms exhibit meaningful interference fringes. We demonstrate that a combination of the DDD camera and sparse coding reduces the requisite dose used to obtain holograms to values less than one-thousandth compared with the CCD camera without image postprocessing. This combination is expected to generate lower-dose and/or higher-speed electron holography.

Ultimate resolution limits of speckle-based compressive imaging

Compressive imaging using sparsity constraints is a very promising field of microscopy that provides a dramatic enhancement of the spatial resolution beyond the Abbe diffraction limit. Moreover, it simultaneously overcomes the Nyquist limit by reconstructing an N-pixel image from less than N single-point measurements. Here we present fundamental resolution limits of noiseless compressive imaging via sparsity constraints, speckle illumination and single-pixel detection. We addressed the experimental setup that uses randomly generated speckle patterns (in a scattering media or a multimode fiber). The optimal number of measurements, the ultimate spatial resolution limit and the surprisingly important role of discretization are demonstrated by the theoretical analysis and numerical simulations. We show that, in contrast to conventional microscopy, oversampling may decrease the resolution and reconstruction quality of compressive imaging.

Simulation-Trained Sparse Coding for High-Precision Phase Imaging in Low-Dose Electron Holography

We broaden the applicability of sparse coding, a machine learning method, to low-dose electron holography by using simulated holograms for learning and validation processes. The holograms, with shot noise, are prepared to generate a model, or a dictionary, that includes basic features representing interference fringes. The dictionary is applied to sparse representations of other simulated holograms with various signal-to-noise ratios (SNRs). Results demonstrate that this approach successfully removes noise for holograms with an extremely small SNR of 0.10, and that the denoised holograms provide the accurate phase distribution. Furthermore, this study demonstrates that the dictionary learned from the simulated holograms can be applied to denoising of experimental holograms of a p-n junction specimen recorded with different exposure times. The results indicate that the simulation-trained sparse coding is suitable for use over a wide range of imaging conditions, in particular for observing electron beam-sensitive materials.

Resolution enhancement in scanning electron microscopy using deep learning

We report resolution enhancement in scanning electron microscopy (SEM) images using a generative adversarial network. We demonstrate the veracity of this deep learning-based super-resolution technique by inferring unresolved features in low-resolution SEM images and comparing them with the accurately co-registered high-resolution SEM images of the same samples. Through spatial frequency analysis, we also report that our method generates images with frequency spectra matching higher resolution SEM images of the same fields-of-view. By using this technique, higher resolution SEM images can be taken faster, while also reducing both electron charging and damage to the samples.