Search

Background: This cross-sectional study aimed to examine the association between dietary diversity and risk of subjective cognitive decline (SCD), a precursor of dementia, in middle-aged and elderly Chinese populations residing in eastern China. Methods: Participants aged ≥ 45 years were recruited from a community in an eastern Chinese city after excluding potential objective cognitive impairment using the Mini-Cognitive Assessment Instrument (Mini-Cog). SCD was assessed using the Subjective Cognitive Decline Questionnaire-9 (SCD-Q9). Dietary data were collected using the Dietary Quality Questionnaire (DQQ), and the Food Group Diversity Score (FGDS) and the Consumed All Five Recommended Food Score (All-5) were calculated as indicators of dietary diversity. Odds ratios (ORs) and 95% confidence intervals (CIs) were computed to evaluate the associations of FGDS and All-5 scores with SCD after adjusting for age, sex, socioeconomic status, lifestyle factors, and health status. Results: Among the 871 participants, 358 (41.1%) were classified as having SCD. Compared with participants with the highest FGDS (≥8) and those with the highest All-5 score (5), those with the lowest FGDS (≤4) and the lowest All-5 score (≤3) exhibited 85% (OR = 1.85; 95% CI: 1.10–3.13; p = 0.02) and 90% (OR = 1.90; 95% CI: 1.21–2.97; p < 0.01) higher risk of SCD, respectively, after adjusting for all covariates. Fruits were the only food group among the All-5 components that demonstrated a significant association with SCD risk. Conclusions: Poor dietary diversity was associated with an elevated risk of SCD in middle-aged and older adults, and fruits were the food group with the most substantial effect. [ABSTRACT FROM AUTHOR]

With the advancement of deep learning (DL) technology, DL-based intrusion detection models have emerged as a focal point of research within the domain of cybersecurity. This paper provides an overview of the datasets frequently utilized in the research. This article presents an overview of the widely utilized datasets in the research, establishing a basis for future investigation and analysis. The text subsequently summarizes the prevalent data preprocessing methods and feature engineering techniques utilized in intrusion detection. Following this, it provides a review of seven deep learning-based intrusion detection models, namely, deep autoencoders, deep belief networks, deep neural networks, convolutional neural networks, recurrent neural networks, generative adversarial networks, and transformers. Each model is examined from various dimensions, highlighting their unique architectures and applications within the context of cybersecurity. Furthermore, this paper broadens its scope to include intrusion detection techniques facilitated by the following two large-scale predictive models: the BERT series and the GPT series. These models, leveraging the power of transformers and attention mechanisms, have demonstrated remarkable capabilities in understanding and processing sequential data. In light of these findings, this paper concludes with a prospective outlook on future research directions. Four key areas have been identified for further research. By addressing these issues and advancing research in the aforementioned areas, this paper envisions a future in which DL-based intrusion detection systems are not only more accurate and efficient but also better aligned with the dynamic and evolving landscape of cybersecurity threats. [ABSTRACT FROM AUTHOR]

Leaf chlorophyll content is an important indicator of photosynthetic capacity and health status and plays a key role in monitoring plant growth. At present, research on chlorophyll content monitoring has mainly focused on establishing a relationship between vegetation index and chlorophyll content, which was composed of several hyperspectral wavelengths, resulting in the loss of effective information contained in certain wavelengths strongly associated with disease. Therefore, it was very important to select spectral wavelengths to reveal the hyperspectral response of chlorophyll content in diseased leaves. Fifty groups of rice leaves infected with bacterial leaf blight were studied, and hyperspectral images and soil plant analysis development (SPAD) values of disease‐stressed rice leaves were collected every 5 days from the first day after inoculation. A method for predicting the SPAD value of diseased rice leaves using random forest (RF) combined with a back‐propagation optimized neural network (BPNN) was proposed. The bird swarm algorithm (BSA) and the immune algorithm (IA) were used to solve the problems of information loss, local minimization, and slow convergence of the BPNN. The BPNN, BSA–BPNN, IA–BPNN, and BSA–IA–BPNN prediction models for the SPAD value of diseased rice leaves are established. The results show that the BSA–IA–BPNN model with characteristic wavelengths selected by RF achieved the best prediction effect, and the prediction sets R2, root‐mean‐square error, and relative error are 0.97%, 4.34%, and 3.33% respectively. The shortest prediction modeling time is 1.54 s, and the model exhibits the best stability. This study provides allegorical support and a theoretical basis for achieving rapid, nondestructive detection of the SPAD value of diseased rice leaves, provides an effective reference for estimating the SPAD value of diseased rice leaves in the field, and is of great importance for rice disease monitoring. [ABSTRACT FROM AUTHOR]

The porosity of the gas diffusion layer (GDL) significantly impacts the performance of proton exchange membrane fuel cells (PEMFCs). Assembly pressure in PEMFCs leads to GDL deformation and alterations in porosity distribution. This study integrated a three‐dimensional (3D) GDL deformation model with a 3D two‐phase PEMFC model, employing a four‐term Fourier series model to optimize the fitting of the GDL porosity distribution curve. The approach quantitatively assessed the impact of GDL porosity distribution under assembly pressure on PEMFC performance. Results reveal an arched porosity distribution in GDL, peaking in the middle of low channels adjacent to ribs. High porosity enhances oxygen and heat conduction but excessive porosity may cause uneven current density distribution, hindering GDL drainage. Furthermore, the analysis compares performances at various GDL compression ratios and thicknesses, showing an initial rise then fall in current density with increasing pressure. This represents a trade‐off between the adverse impact of GDL compression on mass transfer losses and the favorable impact of reduced ohmic losses. At the optimal pressure, the current density is 3% higher than neighboring values at the same potential, and within the optimal GDL thickness range, the current density error remains below 1%. [ABSTRACT FROM AUTHOR]

Forest soil contamination is a significant problem with risks to ecosystems and human health. It reduces soil quality, hampers plant growth, and disrupts ecosystems. To understand current research and identify future directions, this study analyzed 2659 documents on forest soil contamination published on the Web of Science from 1970 to 2023. Using bibliometrics, this study systematically analyzed the knowledge structure, research hotspots, and development trends in forest soil pollution. China, the United States, and Poland were the top contributors, with 11.28%, 8.42%, and 7.15% of publications, respectively. Despite fewer publications, the Netherlands and Sweden had significant research influence. The Chinese Academy of Sciences had the most publications. The primary research topics included heavy metals, ecosystems, deposition, air pollution, and organic matter. Keyword cluster and burst analysis highlighted the importance of heavy metals, microbial communities, atmospheric deposition, and organic matter. Notably, microplastics emerged as a notable gap in the existing research by highly cited papers analysis, indicating they can be a future research focus. Overall, this paper provides a comprehensive analysis of forest soil contamination, offering insights into current research themes and emerging trends. [ABSTRACT FROM AUTHOR]

Saffron (Crocus sativus), as an herbal medicine, has been extensively investigated for treating neurological and psychiatric disorders. This systematic review aimed to assess the overall effects of saffron on cognition, depression, anxiety, sleep disorders, attention‐deficit/hyperactivity disorder (ADHD), and obsessive–compulsive disorder (OCD). Relevant randomized controlled trials (RCTs) were identified by searching PubMed/Medline, Web of Science, and Clinical Trials databases up to June 2023 according to search terms and inclusion criteria. The participants were either healthy or suffering from some diseases, including neurological and psychiatric disorders, and consumed saffron or its extracts as an intervention. The risk of bias was assessed according to the Cochrane guidelines, and the PRISMA statement was followed. The meta‐analysis was performed using RevMan and STATA software. A random‐effects or fixed‐effects model was used to calculate the pooled effect sizes. Forty‐six RCTs were enrolled, and the duration of these trials ranged from 4 to 48 weeks with saffron or its extracts, both alone or in combination with conventional drugs. Saffron was more effective than placebo in improving cognition, depression with an overall effect size of −4.26 (95% CI: −5.76, −2.77), anxiety of −3.75 (95% CI: −5.83, −1.67), and sleep disorders of −1.91 (95% CI: −2.88, −0.93). Saffron was non‐inferior to conventional drugs for treating cognitive disorders, depression, anxiety, ADHD, and OCD, and it exhibited good tolerance with few side effects. Saffron may exert protective roles for neurological and psychiatric disorders and represents a relatively favorable and safe treatment. [ABSTRACT FROM AUTHOR]

A common feature within most corporate income tax systems is that the cost of debt is deductible as an expenditure when calculating taxable profits. An unintended consequence of this tax distortion is the creation of under-capitalized firms – raising default risk in the process. Using a difference-in-differences approach, this paper shows that a reduction in tax discrimination between debt and equity finance leads to better capitalized banks. The paper exploits the exogenous variation in the tax treatment of debt and equity created by the introduction of an Allowance for Corporate Equity (ACE) system in Italy, to identify whether an ACE positively impacts banks' capital structure. The results demonstrate that a move to an unbiased corporate tax environment increases bank capital ratios, driven by an increase in equity rather than a reduction in lending activities. The change also leads to a reduction in risk taking for ex-ante low capitalized banks. Overall, these results suggest that the ACE could be a valuable policy instrument for prudential bank regulators. [ABSTRACT FROM AUTHOR]

Chlorinated paraffins (CPs), which were conventionally classified into short- (SCCPs), medium- (MCCPs) and long- (LCCPs) chain CPs, have received growing attention due to their wide usage and extensive detection in environmental samples and biota. The number of studies regarding the biomonitoring of CPs in human beings increased rapidly and their health risk gained great concern. This review summarized their occurrence and homologue patterns in human matrices including blood/serum, placenta, cord serum and breast milk. As the production and usage of SCCPs was progressively banned after being listed in Annex A of the Stockholm Convention, the production of MCCPs and LCCPs was stimulated. Accordingly, the ratio of MCCPs/SCCPs in human samples has increased rapidly in the last 5 years. The current understanding of exposure routes and risk assessments of CPs was also reviewed. Oral dietary intake is the most predominant source of daily CP intake, but dust ingestion, inhalation and dermal exposure is also nonnegligible, especially for MCCPs and LCCPs. Furthermore, the reported upper bound of the estimated daily intakes (EDIs) in various risk assessment studies was close to or exceeded the tolerable daily intakes (TDIs). Considering the bioaccumulation and long-lasting exposure of CPs, their health impacts on humans and the ecosystem required continuous monitoring and evaluation. [ABSTRACT FROM AUTHOR]

The electronic properties of hydrogen-terminated biphenylene (BP) segments of different sizes on the sub-nanoscale are explored using density functional theory, and the size dependence of the energy gap is evaluated using a structural parameter as a function of the bond lengths and the electronic density contributions. More importantly, the energy gap is observed to decrease linearly with the reduced hydrogen-to-carbon ratio of the corresponding structures, while the decrease-rate undergoes a diminution of four times at a gap of 0.5 eV due to the transformed distribution of the lowest unoccupied molecular orbital. The results give a deep insight into the size-tunable energy gaps of BPs and provide a possibility for the preparation of hydrogen-terminated carbon materials with a desirable energy gap. [ABSTRACT FROM AUTHOR]

Strawberry anthracnose, caused by Colletotrichum spp., is a major disease that causes tremendous damage to cultivated strawberry plants (Fragaria × ananassa Duch.). Examining and distinguishing plants potentially carrying the pathogen is one of the most effective ways to prevent and control strawberry anthracnose disease. Herein, we used this method on Colletotrichum gloeosporioides at the crown site on indoor strawberry plants and established a classification and distinguishing model based on measurement of the spectral and textural characteristics of the disease-free zone near the disease center. The results, based on the successive projection algorithm (SPA), competitive adaptive reweighted sampling (CARS), and interval random frog (IRF), extracted 5, 14, and 11 characteristic wavelengths, respectively. The SPA extracted fewer effective characteristic wavelengths, while IRF covered more information. A total of 12 dimensional texture features (TFs) were extracted from the first three minimum noise fraction (MNF) images using a grayscale co-occurrence matrix (GLCM). The combined dataset modeling of spectral and TFs performed better than single-feature modeling. The accuracy rates of the IRF + TF + BP model test set for healthy, asymptomatic, and symptomatic samples were 99.1%, 93.5%, and 94.5%, the recall rates were 100%, 94%, and 93%, and the F1 scores were 0.9955, 0.9375, and 0.9374, respectively. The total modeling time was 10.9 s, meaning that this model demonstrated the best comprehensive performance of all the constructed models. The model lays a technical foundation for the early, non-destructive detection of strawberry anthracnose. [ABSTRACT FROM AUTHOR]

Background Linezolid-induced thrombocytopenia is the main factor restricting the clinical application of linezolid. Objectives To investigate the relationship between PNU-14230 concentration and linezolid-induced thrombocytopenia and further develop and validate a risk model for predicting linezolid-induced thrombocytopenia. Methods A regression model was constructed to predict the occurrence of linezolid-induced thrombocytopenia, and further externally validated. The predictive performance was evaluated by receiver operating characteristic curve and Hosmer–Lemeshow test. Linezolid C min and PNU-142300 concentrations were compared for different kidney function groups. The Kaplan–Meier method was used to estimate the difference in cumulative incidence of linezolid-induced thrombocytopenia among different kidney function patients. Results In the derivation (n = 221) and validation (n = 158) cohorts, 28.5% and 24.1% of critically ill patients developed linezolid-induced thrombocytopenia. Logistic regression analysis indicated that the independent risk factors were linezolid C min, PNU-142300 concentration, baseline platelet count, renal insufficiency (RI) and continuous venovenous haemofiltration (CVVH). The AUC for the risk model was 0.901, and the model was good (P  = 0.633). The model also showed good discrimination (AUC 0.870) and calibration (P  = 0.282) in the external validation cohort. Compared with normal kidney function patients, patients with RI and CVVH had higher linezolid C min and PNU-142300 concentrations (P < 0.001) and higher cumulative incidence of linezolid-induced thrombocytopenia (P < 0.001). Conclusions PNU142300 concentration, as well as linezolid C min, might identify patients at risk of linezolid-induced thrombocytopenia. The risk prediction model had good predictive performance for linezolid-induced thrombocytopenia development. Concentrations of linezolid and PNU-142300 accumulated in patients with RI and CVVH. [ABSTRACT FROM AUTHOR]

Cross‐modal prediction serves a crucial adaptive role in the multisensory world, yet the neural mechanisms underlying this prediction are poorly understood. The present study addressed this important question by combining a novel audiovisual sequence memory task, functional magnetic resonance imaging (fMRI), and multivariate neural representational analyses. Our behavioral results revealed a reliable asymmetric cross‐modal predictive effect, with a stronger prediction from visual to auditory (VA) modality than auditory to visual (AV) modality. Mirroring the behavioral pattern, we found the superior parietal lobe (SPL) showed higher pattern similarity for VA than AV pairs, and the strength of the predictive coding in the SPL was positively correlated with the behavioral predictive effect in the VA condition. Representational connectivity analyses further revealed that the SPL mediated the neural pathway from the visual to the auditory cortex in the VA condition but was not involved in the auditory to visual cortex pathway in the AV condition. Direct neural pathways within the unimodal regions were found for the visual‐to‐visual and auditory‐to‐auditory predictions. Together, these results provide novel insights into the neural mechanisms underlying cross‐modal sequence prediction. [ABSTRACT FROM AUTHOR]

We test the standard New Keynesian (NK) dynamic stochastic general equilibrium (DSGE) model with and without inattentiveness features, where inattentiveness is modeled either in the form of sticky information (SI) or imperfect information (IF) data revision. All models are estimated and tested by indirect inference. The estimated model with sticky information somewhat outperformed the estimated NK model with full‐information rational expectation (FIRE), though both passed the test easily, while the model with IF data revision was strongly rejected. [ABSTRACT FROM AUTHOR]

Rice disease detection is of great significance to rice disease management. It is difficult to identify the rice leaves with different colors in different disease periods by RGB image and without aided eyes. Traditional equipment and methods are relatively inefficient in meeting the needs of current disease detection. The accurate and efficient detection the infected areas from hyperspectral images has become a primary concern in current research. However, current spectral target detection research pays less attention to the time and computing resources consumed by detection. A disease detection method based on random forest (RF) and adaptive coherence estimator (ACE) is proposed here. Firstly, based on the spectral differences between diseased and healthy leaves, 18 characteristic spectral wavelengths with the highest importance were selected by an RF algorithm, and the spectral images of those characteristic wavelengths were synthesized. Then, the ACE model was established for the disease recognition of full wavelength spectral images, characteristic wavelength spectral images, and RGB images. At the same time, three other familiar target detection methods were selected as the control experiments. The detection results showed a similarity between the detection performance of the four detection methods for full wavelength spectral image and characteristic wavelength spectral image. This detection performance was higher than that of the RGB image, indicating that characteristic wavelength spectral image can replace full wavelength spectral image for disease detection. The detection performance of the ACE algorithm was better than other algorithms. The detection accuracy of 18 characteristic wavelengths was 97.41%. Compared with the hyperspectral full wavelength image detection results, the accuracy decreased by 1.12%, and the detection time decreased by 2/3, which greatly reduced the detection time. Based on these results, the target detection method combining the RF algorithm and the ACE algorithm can effectively and accurately detect rice bacterial blight disease, which provides a new method for automatic detection of plant disease in the field. [ABSTRACT FROM AUTHOR]

Generic visual object tracking is challenging due to various difficulties, e.g. scale variations and deformations. To solve those problems, we propose a novel multi-scale selective kernel module for tracking, which contains small-scale and large-scale branches to model the target at different scales and attention mechanism to capture the more effective appearance information of the target. In our module, we cascade multiple small-scale convolutional blocks as an equivalent large-scale branch to extract large-scale features of the target effectively. Besides, we present a hybrid strategy for feature selection to extract significant information from features of different scales. Based on the current excellent segmentation tracking framework, we propose a novel tracking network that leverages our module at multiple places in the up-sample phase to construct a more accurate and robust appearance model. Extensive experimental results show that our tracker outperforms other state-of-the-art trackers on multiple challenging benchmarks including VOT2018, TrackingNet, DAVIS-2017, and YouTube-VOS-2018 while achieves real-time tracking. [ABSTRACT FROM AUTHOR]