Browse Articles

A robust hydroponic system for horticulture farming using deep learning, IoT, and mobile application

Due to limited literacy among root-level farmers, hydroponic farming in Bangladesh faces significant challenges. Therefore, there is a demand for easy-to-use technical systems to help farmers to monitor and operate smart systems. To address the issue, this study introduces a robust hydroponic system that provides automatic guidelines, monitoring, and a disease detection system. The main objective of this paper is to support farmers by making the cultivation process more convenient and less stressful. The system is structured into three phases: hardware implementation using WeMos controllers, disease detection using the Deep Learning model, and mobile application development for sensor data analysis and automatic notifications. The proposed system significantly demonstrates a high disease detection accuracy of 98.5%. Moreover, the survey report shows that around 80% of the root-level farmers find the system helpful for their cultivation process and increase the usability and monitoring of the system. These findings suggest that the proposed system can substantially improve the operational efficiency and sustainability of hydroponic farming, and it has the potential to enable more effective resource management and disease prevention strategies.

Prevalence of common bacterial STI pathogens and the microscopic diagnostic approach to abnormal vaginal discharge in a tertiary care hospital in Bangkok, Thailand

In Thailand, sexually transmitted infections (STIs) persist as a significant public health issue, notwithstanding the affordability of treatments. The primary challenge lies in diagnostic methodologies. According to the Thai National Treatment Guidelines for abnormal vaginal discharge, wet preparation using proportion of white blood cell (WBC) counts and epithelial cell (EC) guides presumptive STI treatment. This study investigated the prevalence of common STI pathogens in sexually active women presenting with abnormal vaginal discharge and WBC > EC under microscopy; and the cost-minimization analysis of this approach. A cross-sectional study was done during July 2021–March 2023 at the Siriraj Female STI Clinic, Bangkok, Thailand. The eligible participants were non-pregnant Thai women aged 18–50 years with the following conditions; being sexually active in prior one year, presenting with abnormal vaginal discharge, having WBC > EC under microscopy, and no allergy to cefixime and azithromycin which were the presumptive treatment in the study. The endocervical swabs were sent for molecular diagnosis of STI pathogens (polymerase chain reaction; PCR). Cost-minimization analysis comparing two approaches, PCR and wet preparation, was done. From an initial 199 participants, 186 were eligible. The average age was 31.1 ± 9.4 years and their sex debut was at 19.7 ± 3.8 years. Around 10% of them, sex partners had STIs. Prevalent STI pathogens included C. trachomatis(20.4%), N. gonorrhoeae(7.0%), M. genitalium(5.9%) and T. vaginalis(3.8%). Presumptive treatment yielded no severe immediate or delayed adverse effects. Using wet preparation as a primary test with presumptive treatment, cost per one case cured was almost three times lower than that of using PCR as a primary test (1250.0 vs 3454.4 Thai Baht). In summary, a quarter of the sexually-active Thai women with abnormal vaginal discharge and WBC > EC under microscopy had either C. trachomatis or N. gonorrhoeae. The use of wet preparation-guided presumptive STI treatment is practical and cost-saving, compared with the PCR approach. Trial registration: Thai Clinical Trial Registry (TCTR20210702001, 2 July 2021).

Refining the diagnostic approach to latent tuberculosis Infection with Quantiferon gold plus: A retrospective analysis of borderline results

The Quantiferon Gold Plus (QFT) test, a widely used interferon-γ release assay (IGRA), diagnoses latent tuberculosis infection (LTBI) with a positivity threshold of ≥0.35 IU/mL. Results near this cut-off can be challenging to interpret due to variability from immunological, pre-analytical, and technical factors, prompting recommendations for a borderline range to refine diagnosis and reduce overtreatment. This retrospective study analyzed QFT results from 9,944 patients (2019–2023), establishing ranges: < 0.2 IU/mL as negative, 0.2–0.35 IU/mL as borderline negative, 0.35–0.7 IU/mL as borderline positive, and >0.7 IU/mL as positive. Borderline results occurred in 7.6% of patients, particularly in those born in Africa or South America, and in older individuals. Of 64 patients retested, 60.9% reverted to negative, while 17.1% of borderline negatives later converted to positive or borderline positive. Notably, no active TB cases emerged among those who reverted to negative on repeat testing. These findings emphasize the need for cautious interpretation of borderline QFT results, as their link to active TB progression differs from clear results. The study supports repeat testing of borderline cases to enhance LTBI diagnostic accuracy and inform treatment decisions.

Combining greedy and evolutionary algorithms to maximize influence in networks under deterministic linear threshold model

In the paper we consider the well-known Influence Maximization (IM) and Target Set Selection (TSS) problems for Boolean networks under Deterministic Linear Threshold Model (DLTM). The main novelty of our paper is that we state these problems in the context of pseudo-Boolean optimization and solve them using evolutionary algorithms in combination with the known greedy heuristic. We also propose a new variant of (1 + 1)-Evolutionary Algorithm, which is designed to optimize a fitness function on the subset of the Boolean hypercube comprised of vectors of a fixed Hamming weight. The properties of this algorithm suit well for solving IM. The proposed algorithm is combined with the greedy heuristic for solving IM and TSS: the latter is used to construct initial solutions. We show that the described hybrid algorithms demonstrate significantly better performance compared to the computational scheme combining the greedy heuristic with the classic variant of (1 + 1)-EA. In the experiments, the proposed algorithms are applied to both real-world networks and the random networks constructed with respect to well-known models of random graphs. The results show that the new algorithms outperform the competition and are applicable to TSS and IM under DLTM for networks with tens of thousands of vertices.

The arabidopsis WAVE/SCAR protein BRICK1 associates with cell edges and plasmodesmata

Plasmodesmata are specialized structures in plant cell walls that mediate intercellular communication by regulating the trafficking of molecules between adjacent cells. The actin cytoskeleton plays a pivotal role in controlling plasmodesmatal permeability, but the molecular mechanisms underlying this regulation remain unclear. Here, we report that BRK1, a component of the WAVE/SCAR complex involved in Arp2/3-mediated actin nucleation, localizes to PD and primary pit fields in A. thaliana cotyledons, leaves, and hypocotyls. Using a BRK1-YFP reporter line, we detected BRK1 enrichment at cell edges and in primary pit fields, identified by regions of reduced propidium iodide staining. We also observed colocalization between BRK1-YFP and the plasmodesmatal callose stain aniline blue, further supporting BRK1’s association with Plasmodesmata. Together, these findings suggest that the WAVE/SCAR complex participates in plasmodesmatal regulation by promoting ARP2/3-dependent actin filament branching at plasmodesmata, complementing the role of linear actin stabilization by formins.

An enhanced secretary bird optimization algorithm based on precise elimination mechanism and boundary control for numerical optimization and low-light image enhancement

Metaheuristic optimization algorithms often face challenges such as complex modeling, limited adaptability, and a tendency to get trapped in local optima when solving complex optimization problems. To enhance algorithm performance, this paper proposes an enhanced Secretary Bird Optimization Algorithm (MESBOA) based on a precise elimination mechanism and boundary control. The algorithm integrates three key strategies: a precise population elimination strategy, which optimizes the population structure by eliminating individuals with low fitness and intelligently generating new ones; a lens imaging-based opposition learning strategy, which expands the exploration of the solution space through reflection and scaling to reduce the risk of local optima; and a boundary control strategy based on the best individual, which effectively constrains the search range to avoid inefficient searches and premature convergence. Experimental validation shows that on 23 benchmark functions and the CEC2022 test suite, MESBOA significantly outperforms the original Secretary Bird Optimization Algorithm (SBOA) and other comparative algorithms (such as GWO, WOA, PSO, etc.) in terms of convergence speed, solution accuracy, and stability. Taking low-light image enhancement as an application case, MESBOA performs better in metrics such as Mean Squared Error (MSE), Peak Signal-to-Noise Ratio (PSNR), and Structural Similarity Index (SSIM) by optimizing the parameters of the normalized incomplete Beta function, verifying its effectiveness in practical problems. The research indicates that MESBOA provides an efficient solution for complex optimization tasks and has the potential to be promoted and applied in multiple fields.

Comparative analysis of cervical cancer classification of DPAGCHE-enhanced Pap smear images using convolutional neural network models

Cervical cancer remains a significant cause of female mortality worldwide, primarily due to abnormal cell growth in the cervix. This study proposes an automated classification method to enhance detection accuracy and efficiency, addressing contrast and noise issues in traditional diagnostic approaches. The impact of image enhancement on classification performance is evaluated by comparing transfer learning-based Convolutional Neural Network (CNN) models trained on both original and enhanced images. This study employs transfer learning with pre-trained CNNs to classify preprocessed Pap smear images into three categories. Data augmentation, including rotations, flips, and shifts, enhances variability and prevents overfitting. The OneCycle learning rate schedule dynamically adjusts the learning rate, improving training efficiency. To enhance image quality, the Denoised Pairing Adaptive Gamma with Clipping Histogram Equalization (DPAGCHE) method improves contrast and reduces noise. The evaluation involves five pre-trained CNN models and the publicly available Herlev dataset, implemented in MATLAB Online. The ResNet50 model trained on the DPAGCHE-enhanced dataset achieves the highest classification performance, with 84.15% accuracy, along with improved specificity, recall, precision, and F1-score. ResNet50’s residual connections mitigate vanishing gradient issues and enhance deep feature extraction. Accordingly, the DPAGCHE preprocessing significantly improves classification performance, leading to a 53.65% increase in F1-score and 44.29% in precision. In contrast, the Baseline CNN reaches only 66.67% accuracy, highlighting the advantage of deeper architectures combined with enhanced preprocessing. These findings suggest integrating DPAGCHE-enhanced preprocessing with deep learning improves automated cervical cancer detection. In particular, ResNet50 demonstrates the best performance, reinforcing the effectiveness of contrast enhancement and noise reduction in aiding classification models.

Global burden of head and neck cancer from 1990 to 2021: A comprehensive analysis and projections to 2030 based on the global burden of disease study 2021

Background Head and neck cancer (HNC) is a significant global health concern with rising incidence and mortality in certain regions. This study aimed to evaluate the global burden and temporal trends of HNC from 1990 to 2021 and to project its future burden through 2030. Methods Data were obtained from the Global Burden of Disease (GBD) 2021 study. Joinpoint regression was used to assess temporal trends in age-standardized incidence rates (ASIR), age-standardized death rates (ASDR), and disability-adjusted life years (DALYs). Age–period–cohort (APC) and Bayesian APC (BAPC) models were applied to evaluate age, period, and cohort effects and to project future trends. Decomposition analysis was conducted to explore the contributions of population aging, growth, and epidemiological changes. Results In 2021, there were 792,280 new HNC cases and 424,066 deaths globally. Age-standardized incidence rates remained stable, while death rates and DALYs significantly declined. Incidence rose notably in East Asia, whereas mortality and DALYs increased substantially in Oceania. Gender differences were evident, with higher burdens in males, although female incidence rates recently increased. Aging and population growth were key contributors to the rising burden. Projections suggest a notable increase in female incidence and continued decline in male DALYs by 2030. Among HNC subtypes, lip and oral cavity cancers had the highest burden, whereas other pharyngeal cancers showed increasing incidence trends. Conclusions Despite overall declines in mortality and disease burden, the global incidence of HNC remains substantial. Targeted interventions—such as tobacco and alcohol control and widespread HPV vaccination—are essential to mitigate the future burden of HNC.

SeqFusionNet: A hybrid model for sequence-aware and globally integrated acoustic representation

Animals communicate information primarily via their calls, and directly using their vocalizations proves essential for executing species conservation and tracking biodiversity. Conventional visual approaches are frequently limited by distance and surroundings, while call-based monitoring concentrates solely on the animals themselves, proving more effective and straightforward than visual techniques. This paper introduces an animal sound classification model named SeqFusionNet, integrating the sequential encoding of Transformer with the global perception of MLP to achieve robust global feature extraction. Research involved compiling and organizing four common acoustic datasets (pig, bird, urbansound, and marine mammal), with extensive experiments exploring the applicability of vocal features across species and the model’s recognition capabilities. Experimental results validate SeqFusionNet’s efficacy in classifying animal calls: it identifies four pig call types at 95.00% accuracy, nine and six bird categories at 94.52% and 95.24% respectively, fifteen and eleven marine mammal types reaching 96.43% and 97.50% accuracy, while attaining 94.39% accuracy on ten urban sound categories. Comparative analysis shows our method surpasses existing approaches. Beyond matching reference models on UrbanSound8K, SeqFusionNet demonstrates strong robustness and generalization across species. This research offers an expandable, efficient framework for automated bioacoustic monitoring, supporting wildlife preservation, ecological studies, and environmental sound analysis applications.

Correction: Detection and characterization of fungus (Magnaporthe oryzae pathotype Triticum) causing wheat blast disease on rain-fed grown wheat (Triticum aestivum L.) in Zambia

Accuracy of recording linear erosion using an unmanned aerial vehicle (UAV)

Soil erosion is an ongoing environmental problem. To address this issue, calibrated erosion models are used to forecast areas vulnerable to erosion and to determine appropriate preventive measures. Model calibrations are based on erosion data recorded using different techniques such as photogrammetry from an unmanned aerial vehicle (UAV). In this study, the accuracy of the DJI P4 RTK UAV data was estimated for cropland boundary conditions. Ground heights of tilled and untilled arable land and standing water surfaces were determined using aerial surveys and compared to terrestrial surveys conducted on site. The results revealed that untilled soils can be accurately detected using a UAV, whereas the detection error rates of tilled soils were 2–3 folds higher. Additionally, the width and height of linear erosion tracks were measured and compared using aerial surveys and manual on-site measurements. The erosion width of the linear tracks was accurately recorded using a UAV whereas the erosion depth was underestimated by the digital elevation model (DEM) generated from UAV data.

Retraction: Pre-infection 25-hydroxyvitamin D3 levels and association with severity of COVID-19 illness

Tutorial on computing nonadiabatic proton-coupled electron transfer rate constants

Proton-coupled electron transfer (PCET) is pervasive throughout chemistry, biology, and physics. Over the last few decades, we have developed a general theoretical formulation for PCET that includes the quantum mechanical effects of the electrons and transferring protons, including hydrogen tunneling, as well as the reorganization of the environment and the donor–acceptor fluctuations. Analytical rate constants have been derived in various well-defined regimes. This Tutorial focuses on the vibronically nonadiabatic regime, in which a golden rule rate constant expression is applicable. The goal is to provide detailed instructions on how to compute the input quantities to this rate constant expression for PCET in molecules, proteins, and electrochemical systems. The required input quantities are the inner-sphere and outer-sphere reorganization energies, the diabatic proton potential energy profiles, the electronic coupling, the reaction free energy, and the proton donor–acceptor distance distribution function. Instructions on how to determine the degree of electron–proton nonadiabaticity, which is important for determining the form of the vibronic coupling, are also provided. Detailed examples are given for thermal enzymatic PCET, homogeneous molecular electrochemical PCET, photochemical molecular PCET, and heterogeneous electrochemical PCET. A Python-based package, pyPCET, for computing nonadiabatic PCET rate constants, along with example scripts, input data, output files, and detailed documentation, is publicly available.

Understanding the shape of chemistry data—Applications with persistent homology

Chemical data often have complex and nonlinear patterns in how data points relate to one another. Concurrently, there are many situations where chemical data are of high dimensionality (e.g., the 3N-dimensional potential energy landscape). Both complexity and high dimensionality pose challenges for analyses that seek to uncover fundamental structure–property relationships or to develop foundational models of chemical behavior. This Perspective offers mathematical context, illustrative applications, and conceptual motivation for using persistent homology (PH) to identify and provide new physical insight into the multiple spatiotemporal-scale patterns present in chemical data. We address the implications of different data representations and highlight the relationships of PH-derived descriptors to physicochemical properties and chemical behavior. Applications in machine learning are also discussed, emphasizing how PH can enhance predictive modeling. Finally, we review commonly used PH software, offering recommendations on usability, flexibility, and data requirements.

Propargyl (∙C3H3) and butadienyl (∙<i>i</i>-C4H5) radical–radical reactions well-skipping to vinylcyclopentadienyl radical and toluene: A theoretical and kinetic modeling study

Propargyl radical (•C3H3) and butadienyl radical (•i-C4H5) are two crucial intermediates in combustion and astrochemistry, particularly in the formation of C7H8 aromatics such as toluene. However, the precise formation mechanisms of the first-ring aromatics through C3 + C4 reactions have remained ambiguous. This study explores the detailed potential energy surface (PES) of C7H8 at the •C3H3 + •i-C4H5 entrance reaction channel, alongside conducting kinetic calculations and modeling. The PES reveals distinct mechanistic pathways that depend on the resonance configurations of •C3H3 (propyne-3-yl and allenyl-1-yl). Key C7H8 isomers, including 5-ethylidenecyclopenta-1,3-diene, cycloheptatriene, and norcaradiene, are preferentially formed via the allenyl-1-yl configuration, underlining the significant influence of π electron delocalization of propargyl. Kinetic analysis using the phase space theory and the RRKM/ME method identifies well-skipping reactions, leading to larger resonance-stabilized •C7H7 radical and hydrogen atom through the less dominant allenyl-1-yl configuration reacting with •i-C4H5. Rate constants for •C3H3 + •i-C4H5 reaction yielding toluene and vinylcyclopentadienyl (vinylCPDyl) + H are determined. Subsequent kinetic modeling indicates that the formation pathway •C3H3 + •i-C4H5 → toluene predominates at low temperatures and pressure, contrasting with other toluene formations via benzyl + H and phenyl + CH3 reactions. •C3H3 + •i-C4H5 reaction is also notably significant for generating vinylCPDyl at temperatures exceeding 1050 K at 760 Torr. Although polycyclic aromatic hydrocarbons (PAHs) typically form in high-temperature scenarios, this research suggests viable low-temperature pathways for toluene, which are important in cooling zones of engines, thereby influencing PAH and soot production via resonance stabilized radical chain reactions.

Superatomic 1S orbital-mediated ethylene activation on Ag<i>n</i>− clusters

Single-cluster catalysts (SCCs) leverage superatomic properties via well-defined geometric/electronic configurations to enable novel reactions. The development of SCCs has facilitated atomic-level insights into catalyst design, thereby advancing our understanding of the fundamental nature of catalytic reactions. While orbital symmetry rules guide unimolecular catalyst design, the role of superatomic orbital symmetry in SCC reactivity remains elusive. Herein, we systematically investigated the gas-phase reactions of Agn− (n = 7–25) clusters with C2H4, employing a combination of time-of-flight mass spectroscopy and density functional theory calculations. This revealed strong size-dependent reactivity: Ag7–11, 18–23− showed remarkable stability, whereas Ag12–17, 24–25− adsorbed one or even two C2H4 molecules. The electron clouds of 1S superatomic orbital in Ag12–15, 24–25− clusters are partially localized on specific atoms. This partial localization enables effective interactions between the 1S orbital and the π orbital of C2H4, while concurrently enhancing stability through the formation of bonding orbitals and the d-orbital coupling among silver atoms. Notably, C2H4 adsorption induces structural reorganization of Ag16, 17−, resulting in the formation of icosahedral cages. These cages contain highly symmetrical electron clouds that provide symmetrically matched orbitals, favoring secondary C2H4 adsorption and thereby enhancing the stability of complexes. Our research introduces a novel framework for the precision engineering of superatomic clusters while broadening the application scope of the superatomic properties of metal clusters. The discovery of the superatomic orbital symmetry rule sheds light on the activity series of SCCs and offers new insights into precise SCC engineering.

Hexagonal ice density dependence on interatomic distance changes due to nuclear quantum effects

Hexagonal ice (Ih), the most common structure of ice, displays a variety of fascinating properties. Despite major efforts, a theoretical description of all its properties is still lacking. In particular, correctly accounting for its density and interatomic interactions is of utmost importance as a stepping stone for a deeper understanding of other properties. Deep potentials are a recent alternative to investigate the properties of ice Ih, aiming to match the accuracy of ab initio simulations with the simplicity and scalability of classical molecular dynamics. This becomes particularly significant if one wishes to address nuclear quantum effects. In this work, we use machine learning potentials obtained for different exchange and correlation functionals to simulate the structural and vibrational properties of ice Ih. We show that most functionals overestimate the density of ice compared to experimental results. Furthermore, a quantum treatment of the nuclei leads to even further distancing from experiments. We understand this by highlighting how different interatomic interactions play a role in obtaining the equilibrium density. In particular, different from water clusters and bulk water, nuclear quantum effects lead to stronger H-bonds in ice Ih.

Ion association and hydrogen bonding in potassium dihydrogen phosphate solutions: Insights from molecular dynamics simulations

Potassium dihydrogen phosphate (KDP) is a critical material in non-linear optics, with significant applications in electro-optical and laser technologies. Despite its importance, the solution properties of KDP remain poorly understood, and to the best of our knowledge, no prior molecular dynamics (MD) simulation studies have directly probed the structure and behavior of KDP in aqueous solutions. This study presents results from MD simulations of KDP in both solution and solid states and compares four dihydrogen phosphate (DP) force-field models in five different water models. Our results reveal that the solution structure is strongly dominated by the association of the DP anions through direct hydrogen bonding, where the degree of association exhibits a marked concentration dependence in accord with the experiment. Of the four DP models evaluated, two are much better able to reproduce experimental data, including from neutron scattering and ab initio MD simulation results, demonstrating their effectiveness in capturing the hydrogen bonding patterns that appear to govern local solution structure. We find that the extent of hydrogen bonding between DP anions is also sensitive to the choice of water models, with stronger hydration reducing DP–DP association. We have also examined the two models for both tetragonal and monoclinic crystal structures of KDP and found that these models are able to reproduce the experimental parameters relatively well. The findings of this study enhance our understanding of KDP solutions and lay the groundwork for future investigations into its solution and solid-state properties and behavior.

Interfacial degradation of PEO-based polymer electrolytes on the NMC cathode and CEI components prediction

All-solid-state Li-metal batteries using solid polymer electrolytes (SPEs) in combination with high-voltage cathodes such as lithium nickel manganese cobalt oxide (NMC) promise enhanced battery safety, energy density, and flexibility. However, understanding the oxidative decomposition of SPEs on the cathode surfaces and characterizing the resulting cathode-electrolyte interphase (CEI) remain challenging both experimentally and computationally. This study introduces a new computational protocol based on ab initio molecular dynamics for simulating the decomposition of PEO:LiTFSI SPE on the NMC-811 cathode surface using a combined electron- and Li+-removal simulation approach. This method incorporates the effects of the applied electric potential and Li+ migration on electrolyte oxidation during battery charging. The calculations indicate that electrons are withdrawn from both the C–C bonds of PEO and the Ni–O bonds of NMC-811, resulting in C–C bond cleavage and the formation of decomposition fragments. The created Li vacancies in the NMC facilitate coupling between decomposed PEO and exposed surface oxygen. The ROCH2O-M species, identified as the major degradation product on the NMC-811 cathode surface, is in agreement with the experimental XPS spectra. This approach provides detailed insights into the oxidative decomposition of PEO-based SPEs and demonstrates its effectiveness in exploring CEI component formation.

Fast simulation of soft x-ray near-edge spectra using a relativistic state-interaction approach: Application to closed-shell transition metal complexes

Spectroscopic techniques based on core-level excitations offer powerful tools for probing molecular and electronic structures with high spatial resolution. However, accurately calculating spectral features at the L or M edges is challenging due to the significant influence of spin–orbit and multiplet effects. While scalar-relativistic effects can be incorporated with minimal computational cost, accounting for spin–orbit interactions requires complex frameworks that can be computationally expensive. In this work, we develop a reduced-cost state-interaction approach for simulating near-edge soft x-ray absorption spectra of closed-shell transition metal complexes with relativistic effects incorporated using the ZORA-Kohn–Sham Hamiltonian. The computed spectra closely agree with those obtained with state-of-the-art approaches. This methodology provides a practical and cost-effective alternative to more rigorous two-component methods, making it particularly valuable for large-scale calculations and applications such as resonant inelastic x-ray scattering simulations, where capturing a large number of excited states is essential.