Multi-Branch Mutual-Distillation Transformer for EEG-Based Seizure Subtype Classification.

Cross-subject electroencephalogram (EEG) based seizure subtype classification is very important in precise epilepsy diagnostics. Deep learning is a promising solution, due to its ability to automatically extract latent patterns. However, it usually requires a large amount of training data, which may not always be available in clinical practice. This paper proposes Multi-Branch Mutual-Distillation (MBMD) Transformer for cross-subject EEG-based seizure subtype classification, which can be effectively trained from small labeled data. MBMD Transformer replaces all even-numbered encoder blocks of the vanilla Vision Transformer by our designed multi-branch encoder blocks. A mutual-distillation strategy is proposed to transfer knowledge between the raw EEG data and its wavelets of different frequency bands. Experiments on two public EEG datasets demonstrated that our proposed MBMD Transformer outperformed several traditional machine learning and state-of-the-art deep learning approaches. To our knowledge, this is the first work on knowledge distillation for EEG-based seizure subtype classification.

Micro-motion characteristics of multi-feature targets on the double pulse coherent system.

In this paper, we study the micro-motion characteristics of multi-feature targets based on a double pulse coherent system under atmospheric conditions. The theoretical model for echo signal and micro-motion characteristics of a 3D target in double pulse coherent system is deduced. We discuss the influence of micro-motion characteristics, the relative size of light spot and target, target shapes, and incident direction on frequency shift. LRCS (Lidar cross-section), echo waveform, intensity and radiation energy distribution under different conditions are obtained additionally. Simulation results conclude that these parameters are of advantage to the inversion of target shape properties and motion types.

Constructing sulfur and oxygen super-coordinated main-group electrocatalysts for selective and cumulative H2O2 production.

Direct electrosynthesis of hydrogen peroxide (H2O2) via the two-electron oxygen reduction reaction presents a burgeoning alternative to the conventional energy-intensive anthraquinone process for on-site applications. Nevertheless, its adoption is currently hindered by inferior H2O2 selectivity and diminished H2O2 yield induced by consecutive H2O2 reduction or Fenton reactions. Herein, guided by theoretical calculations, we endeavor to overcome this challenge by activating a main-group Pb single-atom catalyst via a local micro-environment engineering strategy employing a sulfur and oxygen super-coordinated structure. The main-group catalyst, synthesized using a carbon dot-assisted pyrolysis technique, displays an industrial current density reaching 400 mA cm-2 and elevated accumulated H2O2 concentrations (1358 mM) with remarkable Faradaic efficiencies. Both experimental results and theoretical simulations elucidate that S and O super-coordination directs a fraction of electrons from the main-group Pb sites to the coordinated oxygen atoms, consequently optimizing the *OOH binding energy and augmenting the 2e- oxygen reduction activity. This work unveils novel avenues for mitigating the production-depletion challenge in H2O2 electrosynthesis through the rational design of main-group catalysts.

Disentangling the role of environmental filtering and biotic resistance on alien invasions in a reservoir area.

Large-scale water conservancy projects benefit human life but have modified the landscape and provided opportunities for alien plant invasions. Understanding the environmental (e.g., climate), human-related (e.g., population density, proximity to human activities), and biotic (e.g., native plant, community structure) factors driving invasions is essential in the management of alien plants and biodiversity conservation in areas with intense human pressure. To this end, we investigated the spatial patterns of alien plant species distribution in the Three Gorges Reservoir Area (TGRA) of China and distinguished the role of the external environment and community characteristics in determining the occurrence of alien plants with differing levels of known invasion impacts in China using random forest analyses and structural equation models. A total of 102 alien plant species belonging to 30 families and 67 genera were recorded, the majority being annual and biennial herbs (65.7%). The results showed a negative diversity-invasibility relationship and supported the biotic resistance hypothesis. Moreover, percentage coverage of native plants was found to interact with native species richness and had a predominant role in resisting alien plant species. We found alien dominance was mainly the result of disturbance (e.g., changes in hydrological regime), which drove native plant loss. Our results also demonstrated that disturbance and temperature were more important for the occurrence of malignant invaders than all alien plants. Overall, our study highlights the importance of restoring diverse and productive native communities in resistance to invasion.

Ergothioneine and its congeners: anti-ageing mechanisms and pharmacophore biosynthesis.

Ergothioneine, Ovothiol, and Selenoneine are sulfur/selenium-containing histidine-derived natural products widely distributed across different organisms. They exhibit significant antioxidant properties, making them as potential lead compounds for promoting health. Increasing evidence suggests that Ergothioneine is positively correlated with healthy ageing and longevity. The mechanisms underlying Ergothioneine's regulation of the ageing process at cellular and molecular levels are beginning to be understood. In this review, we provide an in-depth and extensive coverage of the anti-ageing studies on Ergothioneine and discuss its possible intracellular targeting pathways. In addition, we highlight the recent efforts in elucidating the biosynthetic details for Ergothioneine, Ovothiol, and Selenoneine, with a particular focus on the study of their pharmacophore-forming enzymology.

Impact of Elevational Gradients and Chemical Parameters on Changes in Soil Bacterial Diversity Under Semiarid Mountain Region.

Elevation gradients, often regarded as "natural experiments or laboratories", can be used to study changes in the distribution of microbial diversity related to changes in environmental conditions that typically occur over small geographical scales. We obtained bacterial sequences using MiSeq sequencing and clustered them into operational taxonomic units (OTUs). The total number of reads obtained by the bacterial 16S rRNA sequencing analysis was 1,090,555, with an average of approximately 45,439 reads per sample collected from various elevations. The current study observed inconsistent bacterial diversity patterns in samples from the lowest to highest elevations. 983 OTUs were found common among all the elevations. The most unique OTUs were found in the soil sample from elevation_2, followed by elevation_1. Soil sample collected at elevation_6 had the least unique OTUs. Actinobacteria, Protobacteria, Chloroflexi were found most abundant bacterial phyla in current study. Ammonium nitrogen (NH4+-N), and total phosphate (TP) are the main factors influencing bacterial diversity at elevations_1. pH was the main factor influencing the bacterial diversity at elevations_2, elevation_3 and elevation_4. Our results provide new visions on forming and maintaining soil microbial diversity along an elevational gradient and have implications for microbial responses to environmental change in semiarid mountain ecosystems.

Error analysis and optimization for a full-Stokes division-of-space polarimeter.

A generalized four-channel, full-Stokes division-of-space (DoSP) error propagation model and its version with a reference optical path are presented in this paper, covering all potential error sources such as the main detector noise, intensity fluctuations, and instrument matrix error. Based on the model, a classical division-of-amplitude polarimeter (DoAmP) structure consisting of a partially polarized beam splitter (PPBS), PBS, and wave plates is thoroughly evaluated. By optimizing the PPBS and azimuth of the wave plates, several optimal parameter configurations are identified where the condition number is 1.84, and the maximum wavelength deviation range is limited to (-3.4n m, 3.62 nm), where the degree of polarization and polarized angle errors do not exceed 0.03 and 0.3°, respectively, and the instrument matrix deterioration effect is minimal enough to be disregarded. In addition to the DoAmP structure, this error propagation model can be directly extended to other arbitrary four-channel DoSP structures such as division-of-focal-plane and division-of-aperture systems, which have guidance values for system structural design, error optimization, and discovering multi-wavelength compatibility of the instrument.

Anthropogenic activities explained the difference in exotic plants invasion between protected and non-protected areas at a northern subtropics biodiversity hotspot.

Biological invasion poses a major threat to biodiversity and conservation efforts in protected areas. The Greater Shennongjia Area (GSA) is one of China's 16 key areas for biodiversity, as stated in the China National Biodiversity Conservation Strategy and Action Plan. However, the local authorities lack appropriate data on the extent and impact of exotic species in protected areas, as well as lack the capacity and motivation to properly plan for exotic species strategy and action plan to support both prevention, control as well as management of exotic plants in their jurisdiction. In addition, while most previous studies have focused on exotic species in protected areas, little effort has been devoted to specifying which environmental factors contribute to the difference between protected and non-protected areas. Here, we explored the current distribution pattern of the richness and abundance of exotic species in relation to environmental variables within the GSA. In total, we found 84 exotic plant species, of which 41 exotic species within the protected areas, in 64 genera and 27 families, predominately from Asteraceae, Fabaceae, and Poaceae. The generalized linear mixed models (GLMMs) revealed that the protection status and the distance to human settlements were the most important predictors of exotic plant richness and abundance in the GSA. Our results showed that the average exotic plant richness and coverage in the protected areas were 22% and 31% lower than outside the protected areas, respectively. Such differences were probably the result of anthropogenic activities (e.g., proximity to human settlements and the proportion of cropland). Although protected areas provide an important barrier against plant invasions, invasion may be a tricky issue for protected area management in the future. The Alliance of Protected areas in Western Hubei and Eastern Chongqing will need to further consider stringent control and management strategies for the entry of exotic species into protected areas to effectively maintain the continuity and integrity of the GSA's biodiversity and ecosystems. Our results provided guidance and support to enhance the capacity of scientific and effective management and sustainable development of the Shennongjia World Natural Heritage Site and other protected areas.

Understanding oxygen evolution mechanisms by tracking charge flow at the atomic level.

Current classifications of oxygen evolution catalysts are based on energy levels of the clean catalysts. It is generally asserted that a LOM-catalyst can only follow LOM chemistry in each electron transfer step and that there can be no mixing between AEM and LOM steps without an external trigger. We use ab initio theory to track the charge flow of the water-on-catalyst system and show that the position of water orbitals is pivotal in determining whether an electron transfer step is water dominated oxidation (WDO), lattice-oxygen dominated oxidation (LoDO), or metal dominated oxidation (MDO). Microscopic photo-catalytic pathways of TiO2 (110), a material whose lattice oxygen bands lie above the metal bands, show that viable OER pathways follow either all AEM steps or mixed AEM-LOM steps. The results provide a correct description of redox chemistries at the atomic level and advance our understanding of how water-splitting catalysts produce desorbed oxygen.

Research on echo characteristics in remote detection with the pulse LiDAR of aerial targets under diverse atmospheric conditions.

This paper utilizes the idea of theoretical analysis to introduce a method for echo characteristics in remote detection with the pulse LiDAR of aerial targets under atmospheric conditions. A Missiles and an aircraft are selected as simulation targets. The relation among mutual mapping of target surface elements can be directly gained by setting light source and target parameters. We discuss influences on atmospheric transport conditions, target shapes and detection conditions on echo characteristics. Atmospheric transport model is introduced as weather conditions, including sunny and cloudy days, with or without turbulence. Simulation results conclude that the outline of scanned waveform can invert the target shape. These provide a theoretical basis for improving the target detection and tracking performance.

Source-Free Domain Adaptation (SFDA) for Privacy-Preserving Seizure Subtype Classification.

Electroencephalogram (EEG) based seizure subtype classification is very important in clinical diagnostics. Source-free domain adaptation (SFDA) uses a pre-trained source model, instead of the source data, for privacy-preserving transfer learning. SFDA is useful in seizure subtype classification, which can protect the privacy of the source patients, while reducing the amount of labeled calibration data for a new patient. This paper introduces semi-supervised transfer boosting (SS-TrBoosting), a boosting-based SFDA approach for seizure subtype classification. We further extend it to unsupervised transfer boosting (U-TrBoosting) for unsupervised SFDA, i.e., the new patient does not need any labeled EEG data. Experiments on three public seizure datasets demonstrated that SS-TrBoosting and U-TrBoosting outperformed multiple classical and state-of-the-art machine learning approaches in cross-dataset/cross-patient seizure subtype classification.

Complex structured beam direct generation by coherent superposition of a complete set of degenerate eigenmodes.

Structured beams have played an important role in many fields due to their rich spatial characteristics. The microchip cavity with a large Fresnel number can directly generate structured beams with complex spatial intensity distribution, which provides convenience for further exploring the formation mechanism of structured beams and realizing low-cost applications. In this article, theoretical and experimental studies are carried out on complex structured beams directly generated by the microchip cavity. It is demonstrated that the complex beams generated by the microchip cavity can be expressed by the coherent superposition of whole transverse eigenmodes within the same order, thus forming the eigenmode spectrum. The mode component analysis of complex propagation-invariant structured beams can be realized by the degenerate eigenmode spectral analysis described in this article.

Universal understanding of self-healing and transformation of complex structured beams based on eigenmode superposition.

The self-healing property of laser beams with special spatial structures is of great interest. We take the Hermite-Gaussian (HG) eigenmode as an example, theoretically and experimentally investigating the self-healing and transformation characteristics of complex structured beams composed of incoherent or coherent superposition of multiple eigenmodes. It is found that a partially blocked single HG mode can recover the original structure or transfer to a lower order distribution in the far field. When the obstacle retains one pair of edged bright spots of the HG mode in each direction of two symmetry axes, the beam structure information (number of knot lines) along each axis can be restored. Otherwise, it will transfer to the corresponding low-order mode or multi-interference fringes in the far field, according to the interval of the two most-edged remaining spots. It is proved that the above effect is induced by the diffraction and interference results of the partially retained light field. This principle is also applicable to other scale-invariant structured beams such as Laguerre-Gauss (LG) beams. The self-healing and transformation characteristics of multi-eigenmode composed beams with specially customized structures can be intuitively investigated based on eigenmode superposition theory. It is found that the HG mode incoherently composed structured beams have a stronger ability to recover themselves in the far field after occlusion. These investigations can expand the applications of optical lattice structures of laser communication, atom optical capture, and optical imaging.

NiO Matrix Decorated by Ru Single Atoms: Electron-Rich Ru-Induced High Activity and Selectivity toward Electrochemical N2 Reduction.

Developing a single-atom catalyst with electron-rich active sites is a promising strategy for catalyzing the electrochemical N2 reduction reaction (NRR). Herein, we choose NiO(001) as a model template and deposit a series of single transition metal (TM) atoms with higher formal charges to create the electron-rich active centers. Our first-principles calculations show that low-valent Ru (+2) on NiO(001) can significantly activate N2, with its oxidation states varying from +2 to +4 throughout the catalytic cycle. The Ru/NiO(001) catalyst exhibits the best activity with a relatively low limiting potential of -0.49 V. Furthermore, under NRR operating conditions, the Ru site is primarily occupied by *N2 rather than *H, indicating that NRR overwhelms the hydrogen evolution reaction and thus exhibits excellent selectivity. Our work highlights the potential of designing catalysts with electron-rich active sites for NRR.

BoostTree and BoostForest for Ensemble Learning.

Bootstrap aggregating (Bagging) and boosting are two popular ensemble learning approaches, which combine multiple base learners to generate a composite model for more accurate and more reliable performance. They have been widely used in biology, engineering, healthcare, etc. This article proposes BoostForest, which is an ensemble learning approach using BoostTree as base learners and can be used for both classification and regression. BoostTree constructs a tree model by gradient boosting. It increases the randomness (diversity) by drawing the cut-points randomly at node splitting. BoostForest further increases the randomness by bootstrapping the training data in constructing different BoostTrees. BoostForest generally outperformed four classical ensemble learning approaches (Random Forest, Extra-Trees, XGBoost and LightGBM) on 35 classification and regression datasets. Remarkably, BoostForest tunes its parameters by simply sampling them randomly from a parameter pool, which can be easily specified, and its ensemble learning framework can also be used to combine many other base learners.

The association between statin use and prognosis in esophageal cancer patients: A meta-analysis.

BACKGROUND: The impact of statin use on the survival of esophageal cancer patients remains unclear now. The aim of this study was to identify the relationship between statin use and the long-term survival of esophageal cancer patients. METHODS: The PubMed, EMBASE, and Web of Science databases were searched up to August 20, 2022, for relevant studies. The endpoints included overall survival (OS), cancer-specific survival (CSS), recurrence-free survival, and hazard ratios (HRs) with corresponding 95% confidence intervals (CIs) were pooled to assess the association between statin use and the prognosis of esophageal cancer patients. Subgroup analysis based on the pathological type (adenocarcinoma vs squamous cell carcinoma), dose of statin use and tumor stage (tumor-node-metastasis I-III vs IV) was further performed. All statistical analyses were conducted using STATA 12.0 software. RESULTS: A total of 7 retrospective studies involving 25,711 participants were included in this meta-analysis. The pooled results indicated that statin use was significantly associated with improved OS (HR = 0.80, 95% CI: 0.74-0.87, P < .001), CSS (HR = 0.77, 95% CI: 0.74-0.89, P < .001), and recurrence-free survival (HR = 0.38, 95% CI: 0.16-0.87, P = .022). Furthermore, subgroup analysis stratified by the pathological type, dose of statin use and tumor stage for OS and CSS showed similar results and indicated the protective role of statin use in the prognosis of esophageal cancer patients. CONCLUSION: Statin use is significantly associated with improved long-term survival of esophageal cancer patients and might serve as a promising prognostic indicator in esophageal cancer. However, more prospective high-quality studies are still needed to verify our findings.

Efficient 38.8 W/m2 solar pumped laser with a Ce:Nd:YAG crystal and a Fresnel lens.

Herein, we report a significant improvement in solar-pumped laser collection efficiency based on end-side pumping a 6-mm-diameter 95-mm-length Ce:Nd:YAG/YAG grooved bonded crystal rod. A Fresnel lens, quartz cooling-water tube, and gold-plated conical cavity constituted the solar-energy collection and concentration system, which was designed to maximum pump light absorption and minimize thermal effects in the Ce:Nd:YAG laser medium. To the best of our knowledge, this is the first time that a Ce:Nd:YAG crystal has been pumped by a Fresnel-lens solar-energy collection and concentration system. The 0.69-m2 effective solar-collection area produced 26.93 W of continuous-wave laser power, corresponding to 6.33% slope efficiency. The collection efficiency (38.8 W/m2) was 1.21 times higher than the highest previously reported value for Fresnel-lens solar collection, and is a record for single-beam solar-pumped lasers.

Activation of CO2 by Direct Cleavage Triggered by Photoelectrons on Rutile TiO2(110).

The initial activation of the inert CO2 is a key step in its photoreduction to valuable chemicals. This process was proposed to proceed mainly by CO2 accepting a photoelectron to form a CO2•- radical or by CO2 accepting two photoelectrons and a proton to form the HCOO- anion on the prototypical rutile TiO2(110) surface. Here, we reveal a new mechanism, in which CO2 is directly cleaved to CO and the adsorbed O2- anion under the trigger of two photoelectrons, by using density functional theory calculations with the HSE06 hybrid functional. The newly revealed mechanism is more favorable than the two previously proposed pathways. Furthermore, our results show that the deficiency of photoelectrons on the catalyst surface is a potential reason for the current low efficiency of CO2 photoreduction.

The surface charge induced high activity of oxygen reduction reaction on the PdTe2 bilayer.

Developing transition metal dichalcogenides as electrocatalysts has attracted great interest due to their tunable electronic properties and good thermal stabilities. Herein, we propose a PdTe2 bilayer as a promising electrocatalyst candidate towards the oxygen reduction reaction (ORR), based on extensive investigation of the electronic properties of PdTe2 thin films as well as atomic-level reaction kinetics at explicit electrode potentials. We verify that under electrochemical reducing conditions, the electron emerging on the electrode surface is directly transferred to O2 adsorbed on the PdTe2 bilayer, which greatly reduces the dissociation barrier of O2, and thereby facilitates the ORR to proceed via a dissociative pathway. Moreover, the barriers of the electrochemical steps in this pathway are all found to be less than 0.1 eV at the ORR limiting potential, demonstrating fast ORR kinetics at ambient conditions. This unique mechanism offers excellent energy efficiency and four-electron selectivity for the PdTe2 bilayer, and it is identified as a promising candidate for fuel cell applications.