Search Results (62,531)

This study addresses the need for efficient and non-destructive monitoring of the nutrient status of hardy kiwi (Actinidia arguta), a plantation crop native to East Asia. Traditional nutrient monitoring methods are labor-intensive and often destructive, limiting their practicality in precision agriculture. To overcome these challenges, we deployed a rotary-wing unmanned aerial vehicle (UAV) equipped with a multispectral camera to capture monthly images of 10 hardy kiwi orchards in South Korea from June to October 2019. We extracted spectral bands (i.e., red, red-edge, green, and near-infrared) to generate normalized difference vegetation index and canopy chlorophyll content index maps, which were correlated with in situ chlorophyll measurements using a chlorophyll meter. Strong positive correlations were observed between vegetation indexes and actual chlorophyll content, with canopy chlorophyll content index achieving the highest predictive accuracy (average correlation coefficient > 0.84). Regression models based on multispectral data enabled reliable estimation of leaf chlorophyll across months and regions, with an average RMSE of 3.1. Our results confirmed that UAV-based multispectral imaging is an effective, scalable approach for real-time monitoring of nutrient status, supporting timely, site-specific fertilizer management. This method has the potential to enhance fertilizer efficiency, reduce environmental impact, and improve the quality of hardy kiwi cultivations. Full article

This study aimed to determine the net energy (NE) values of common energy-supplying nutrients, including starch, protein, and fat, to investigate their influence on energetic efficiency and NE partition patterns in growing pigs, and to develop prediction equations for the protein deposition (PD) and lipid deposition (LD) based on nutrient characteristics of ingredients. Two experiments were conducted. In Experiment 1, 36 growing barrows (Duroc × Landrace × Yorkshire, initial body weight = 28.1 ± 0.8 kg) were randomly allotted to six treatments, with six replicated pigs per treatment. The diets were formulated as follows: a corn–soybean meal basal diet (T1), and five experimental diets containing of 27% corn starch (T2), 27% tapioca starch (T3), 27% pea starch (T4), 5% soybean oil (T5), and 11.8% casein (T6), respectively. In Experiment 2, PD and LD data of 47 ingredients were collected. Subsequently, the nutrient characteristics of ingredients were used as input variables, and PD and LD were used as output variables to establish the prediction equations. Results exhibited that pigs fed the T2, T3, and T4 diets showed increased digestibility of gross energy (GE) and organic matter (OM) compared to those fed the T1 diet (p < 0.01). For various kind of starches, a greater efficiency of using metabolizable energy (ME) for net energy not deposited as protein (PD-free NE, efficiency denoted as kj) was observed when pigs were fed the T2 or T3 diets compared to the T4 diet. Moreover, the kj of soybean oil was 11% and 27% greater than that of starch and casein, respectively, while casein demonstrated 46% and 39% greater efficiency of using ME for PD (efficiency denoted as pj) compared to starch and soybean oil, respectively. Finally, the best-fitted prediction equations for PD and LD were PD = 364.36 − 18.44 × GE + 29.10 × CP − 3.79 × EE − 21.37 × ADF (R² = 0.96; RMSE = 105.15) and LD = −1503.50 + 21.58 × CP + 51.98 × EE + 26.30 × Starch + 26.81 × NDF − 23.87 × ADF (R² = 0.98; RMSE = 172.85), respectively. In summary, there are considerable differences in energetic efficiency and NE partition patterns among various nutrients. In addition, PD and LD can be predicted through nutrient characteristics of ingredients, presenting an innovative approach and methodological framework for the precision nutrition of pigs. Full article

This study proposes a multi-task learning-based production capacity prediction model aimed at improving the prediction accuracy for gas wells in multi-layer tight sandstone reservoirs of the Sulige gas field under small-sample conditions. The model integrates mutation theory and progressive hierarchical feature extraction to achieve adaptive nonlinear feature extraction and autonomous feature selection tailored to different prediction tasks. Using the daily average production of each gas-bearing layer during the first month after well commencement and the cumulative production of each gas-bearing layer over the first year as targets, the model was applied to predict the production capacity of 66 gas wells. Compared with single-task models and classical machine learning methods, the proposed multi-task model significantly improves prediction accuracy, reducing the root mean squared error (RMSE) by over 40% and increasing the coefficient of determination (R²) to 0.82. Experimental results demonstrate the model’s effectiveness in environments with limited training data, offering a reliable approach for productivity prediction in complex multi-layer tight sandstone reservoirs. Full article

Stereotactic radiosurgery (SRS) for multiple brain metastases can be delivered with a single isocenter and non-coplanar arcs, achieving highly conformal dose distributions at the cost of extreme modulation of treatment machine parameters. As a result, SRS plans are at a higher risk of patient-specific quality assurance (PSQA) failure compared to standard treatments. This study aimed to develop a machine-learning (ML) model to predict the PSQA outcome (gamma passing rate, GPR) of SRS plans. Five hundred and ninety-two consecutive patients treated between 2020 and 2024 were selected. GPR analyses were performed using a 3%/1 mm criterion and a 95% action limit for each arc. Fifteen plan complexity metrics were used as input features to predict the GPR of an arc. A stratified and a time-series approach were employed to split the data into training (1555 arcs), validation (389 arcs), and test (486 arcs) sets. The ML model achieved a mean absolute error of 2.6% on the test set, with a 0.83% median residual value (measured/predicted). Lower values of the measured GPR tended to be overestimated. Sensitivity and specificity were 93% and 56%, respectively. ML models for virtual QA of SRS can be integrated into clinical practice, facilitating more efficient PSQA approaches. Full article

Purpose: This study aimed to develop and validate a new machine learning (ML)-based screening tool for a two-step prediction of the need for and type of nutritional therapy (enteral, parenteral, or combined) using Nutrition Risk Screening 2002 (NRS-2002) and other demographic parameters from the Optimal Nutrition Care for All (ONCA) national cohort data. Methods: This multicenter retrospective cohort study included 191,028 patients, with data on age, gender, body mass index (BMI), NRS-2002 score, presence of cancer, and hospital unit type. In the first step, classification models estimated whether patients required nutritional therapy, while the second step predicted the type of therapy. The dataset was divided into 60% training, 20% validation, and 20% test sets. Random Forest (RF), Artificial Neural Network (ANN), deep learning (DL), Elastic Net (EN), and Naive Bayes (NB) algorithms were used for classification. Performance was evaluated using AUC, accuracy, balanced accuracy, MCC, sensitivity, specificity, PPV, NPV, and F₁-score. Results: Of the patients, 54.6% were male, 9.2% had cancer, and 49.9% were hospitalized in internal medicine units. According to NRS-2002, 11.6% were at risk of malnutrition (≥3 points). The DL algorithm performed best in both classification steps. The top three variables for determining the need for nutritional therapy were severe illness, reduced dietary intake in the last week, and mild impaired nutritional status (AUC = 0.933). For determining the type of nutritional therapy, the most important variables were severe illness, severely impaired nutritional status, and ICU admission (AUC = 0.741). Adding gender, cancer status, and ward type to NRS-2002 improved AUC by 0.6% and 3.27% for steps 1 and 2, respectively. Conclusions: Incorporating gender, cancer status, and ward type into the widely used and validated NRS-2002 led to the development of a new scale that accurately classifies nutritional therapy type. This ML-enhanced model has the potential to be integrated into clinical workflows as a decision support system to guide nutritional therapy, although further external validation with larger multinational cohorts is needed. Full article

Groundwater depletion poses a critical challenge to global water security, threatening ecosystems, agriculture, and sustainable development. The Mississippi Delta, a region heavily reliant on groundwater for agriculture, has experienced significant groundwater level declines due to intensive irrigation. Traditional in situ monitoring methods, while valuable, lack the spatial coverage necessary to capture regional groundwater dynamics comprehensively. This study addresses these limitations by leveraging downscaled Gravity Recovery and Climate Experiment (GRACE) data to estimate groundwater levels using random forest modeling (RFM). We applied a machine-learning approach, utilizing the “Forest-based and Boosted Classification and Regression” tool in ArcGIS Pro, (ESRI, Redlands, CA) to predict groundwater levels for April and October over a 10-year period. The model was trained and validated with well-water level records from over 400 monitoring wells, incorporating input variables such as NDVI, temperature, precipitation, and NLDAS data. Cross-validation results demonstrate the model’s high accuracy, with R² values confirming its robustness and reliability. The outputs reveal significant groundwater depletion in the central Mississippi Delta, with the lowest water level observed in the eastern Sunflower and western Leflore Counties. Notably, April 2014 recorded a minimum water level of 18.6 m, while October 2018 showed the lowest post-irrigation water level at 54.9 m. By integrating satellite data with machine learning, this research provides a framework for addressing regional water management challenges and advancing sustainable practices in water-stressed agricultural regions. Full article

Accurate prediction of water quality involves early identification of future pollutant concentrations and water quality indicators, which is an important prerequisite for optimizing water environment management. Although deep learning algorithms have demonstrated considerable potential in predicting water quality parameters, their broader adoption remains hindered by limited interpretability. This study proposes an interpretable deep learning framework integrating an artificial neural network (ANN) model with Shapley additive explanations (SHAP) analysis to predict spatiotemporal variations in water quality and identify key influencing factors. A case study was conducted in the Poyang Lake Basin, utilizing multi-dimensional datasets encompassing topographic, meteorological, socioeconomic, and land use variables. Results indicated that the ANN model exhibited strong predictive performance for dissolved oxygen (DO), total nitrogen (TN), total phosphorus (TP), permanganate index (CODMn), ammonia nitrogen (NH₃N), and turbidity (Turb), achieving R² values ranging from 0.47 to 0.77. Incorporating land use and socioeconomic factors enhanced prediction accuracy by 37.8–246.7% compared to models using only meteorological data. SHAP analysis revealed differences in the dominant factors influencing various water quality parameters. Specifically, cropland area, forest cover, air temperature, and slope in each sub-basin were identified as the most important variables affecting water quality parameters in the case area. These findings provide scientific support for the intelligent management of the regional water environment. Full article

Ferroptosis, a form of cell death, is characterized by lipid peroxidation and is dependent on iron and reactive oxygen species (ROS). Here, through bioinformatics analysis, formosanin C was predicted to be a ferroptosis inducer in colorectal cancer (CRC) by suppressing antioxidation capacity. Indeed, formosanin C induced iron accumulation, lipid ROS formation, and ferroptosis in CRC. We found that TP53 and KRAS were the second and third most frequently mutated genes in CRC and were associated with a poor prognosis. Analyses of differentially expressed genes indicated that fatty acid and labile iron levels tended to be higher in CRC than in normal tissues, suggesting the predisposition of CRC cells to ferroptosis. Transcriptomic analyses in CRC patients further identified that wild-type TP53 and mutant KRAS separately favored ferroptosis. Likewise, p53 knockdown rendered HCT 116 cells less sensitive to ferroptosis, and KRAS HT-29 cells were more sensitive to ferroptosis compared with their parental counterparts. Moreover, formosanin C synergistically enhanced chemosensitivity to cisplatin, and this process was mediated by lipid ROS. Overall, our novel gene-expression screening platform allows for the efficient identification of the biological function of novel phytochemicals, and the data suggest that formosanin C is an effective ferroptosis inducer in CRC cells with p53 or oncogenic KRAS. Full article

This study aims to investigate the molecular mechanisms underlying germ cell tumors (GCTs), focusing specifically on seminomas and teratomas. By analyzing gene expression profiles and miRNA interactions, the goal is to identify key regulatory miRNAs and signaling pathways that differentiate these tumor types and could serve as important regulators for therapy development. Raw data for seminomas and teratomas were extracted from the GEO database, and gene hubs were identified using STRING and Gephi. Signaling pathways and functional annotations were analyzed using miRPathDB, while miRNA–gene interactions were explored via miRWalk. Hub miRNAs were filtered and confirmed using miRDB. This study highlights significant changes in gene expression diversity between tumor and normal gonadal tissues, providing insights into the molecular dynamics of seminomas and teratomas. Distinctions between seminomas and teratomas were identified, shifting the focus toward miRNAs to discover more precise and novel therapeutic approaches. The hub genes of seminomas and teratomas were identified separately. MiRNAs targeting these hub genes were also determined and confirmed. These miRNAs collectively influence essential oncogenic pathways—confirming hsa-miR-138-5p as a regulator of pathways such as Hippo signaling, transcriptional misregulation in cancer, and microRNA cancer signaling in seminomas, and hsa-miR-200b-3p as a regulator of p53 signaling, T cell receptor signaling, and pathways including PI3K/AKT, MAPK/ERK, and Wnt/β-catenin in teratomas—confirming their potential as promising candidates for subtype-specific therapeutic intervention. MiRNAs identified through bioinformatics analyses, and their predicted regulatory roles in key oncogenic pathways, represent potential therapeutic targets or regulators of biological processes. However, further experimental validation is needed to confirm these findings. Full article

Background: Parapneumonic effusion is a common complication of community-acquired pneumonia and can range from a simple inflammatory transudate to an organized purulent collection, known as empyema. Progression to empyema significantly worsens the prognosis, leading to increased morbidity, longer hospital stays, and a greater need for invasive interventions. Several risk factors for pleural effusion and progression to empyema have been identified, but the absence of standardized criteria underline the need for better risk stratification. We analyzed clinical and laboratory data from a cohort of children hospitalized with pneumonia associated with pleural effusion or empyema, to identify predictive risk factors associated with these complications. Methods: We retrospectively analyzed clinical and laboratory data from patients admitted to our Pediatric Emergency Department with pneumonia complicated by pleural effusion and compared patients with simple effusion to those with empyema. Results: Seventeen children with simple pleural effusion and eighteen with empyema were enrolled. Patients with empyema had higher absolute neutrophil count, higher levels of C-reactive protein, procalcitonin, and ferritin, and lower serum albumin levels. Furthermore, they took a longer time for normalization of inflammatory markers when compared with those with pleural effusion. Invasive interventions, such as pleural drainage, and the need for intensive care were more frequent in the empyema group. Conclusions: Pleural effusion and empyema are two common complications of pediatric community-acquired pneumonia. Children developing pleural empyema have higher inflammatory markers and lower levels of serum albumin compared to patients with simple pleural effusion. Morbidity is significantly worse in children with empyema as they are more prone to require invasive interventions and intensive care. Full article

Young children’s problematic media use (PMU) is a growing concern, and parents are critical in shaping early digital habits. However, research often overlooks the dyadic interplay between mothers’ and fathers’ attitudes and parenting practices. This study examined how parents’ favorable attitudes toward child screen media (PASU) predict their own (actor) and their partner’s (partner) mediation behaviors, and how these behaviors subsequently mediate the path to children’s PMU. Drawing on survey data from 1802 matched urban Chinese mother–father pairs, we employed an Actor–Partner Interdependence Mediation Model (APIMeM) within a structural equation modeling (SEM) framework. This dyadic model simultaneously tested actor, partner, and indirect mediation paths connecting parental attitudes to PMU via eight specific parenting practices. Results showed that more positive PASUs predicted each parent’s own supportive behaviors (e.g., high-quality dialogue, autonomy support) but not restrictive limits. Partner effects were modest and asymmetric: mothers’ positive attitudes predicted greater knowledge in fathers, whereas fathers’ positive attitudes were linked to lower communication quality from mothers. Of all parenting dimensions, only higher communication quality (both parents) and mothers’ hands-on monitoring directly predicted lower PMU. Mediation analyses confirmed communication quality as the sole reliable pathway: each parent’s favorable attitudes indirectly lowered PMU by enhancing their own dialogue, but fathers’ attitudes simultaneously increased PMU by eroding mothers’ dialogue. These findings spotlight constructive conversation and coordinated dyadic strategies—especially safeguarding maternal dialogue—as critical targets for interventions aimed at curbing early PMU. Full article

High-temperature heat pumps (HTHPs) have gradually begun to play an essential role in using heat in industry for waste heat recovery and providing higher-grade heat. The isentropic efficiency and volumetric efficiency of HTHPs are significantly affected by high-temperature operating conditions, which take the pressure ratio (PR) as the key parameter, with limited consideration of other factors such as temperature. Relying on the experimental data obtained from the industrial-grade HTHP system experimental platform, this work proposed an XGBoost-MLR algorithm-based method to identify the key parameters of HTHP isentropic efficiency and volumetric efficiency. High-precision (R² > 0.95) prediction models were established to determine the effect of temperature variables on isentropic efficiency and volumetric efficiency. After the key parameters were identified, the empirical equation of isentropic efficiency and volumetric efficiency applicable to this operation condition were constructed. The average relative errors of the two empirical formulas were 5.95% and 5.28%, respectively. Finally, the generalizability of empirical formulas was verified using experimental data from other researchers. The isentropic empirical formula had a relative deviation of less than 10% under twin-screw compressor conditions. However, the applicability of the volumetric efficiency empirical formula was unstable in compressors of different sizes. The feasibility of the method was also discussed. Full article

The rock strata traversed by frozen shafts in coal mines located in western regions are predominantly composed of weakly cemented, water-rich sandstones of the Cretaceous system. Investigating the rheological damage behavior of saturated sandstone under frozen conditions is essential for evaluating the safety and stability of these frozen shafts. To explore the damage evolution and creep characteristics of Cretaceous sandstone under the coupled influence of low temperature and in situ stress, a series of triaxial creep tests were conducted at a constant temperature of −10 °C, under varying confining pressures (0, 2, 4, and 6 MPa). Simultaneously, acoustic emission (AE) energy monitoring was employed to characterize the damage behavior of saturated frozen sandstone under stepwise loading conditions. Based on the experimental findings, a fractional-order creep constitutive model incorporating damage evolution was developed to capture the time-dependent deformation behavior. The sensitivity of model parameters to temperature and confining pressure was also analyzed. The main findings are as follows: (1) Creep deformation progressively increases with higher confining pressure, and nonlinear accelerated creep is observed during the final loading stage. (2) A fractional-order nonlinear creep model accounting for the coupled effects of low temperature, stress, and damage was successfully established based on the test data. (3) Model parameters were identified using the least squares fitting method across different temperature and pressure conditions. The predicted curves closely match the experimental results, validating the accuracy and applicability of the proposed model. These findings provide a theoretical foundation for understanding deformation mechanisms and ensuring the structural integrity of frozen shafts in Cretaceous sandstone formations of western coal mines. Full article

The Third Pole region, particularly the Hindu–Kush–Himalaya (HKH), is highly prone to lightning, causing thousands of fatalities annually. Skillful prediction and timely communication are essential for mitigating lightning-related losses in such observationally data-sparse regions. Therefore, this study evaluates kilometer-scale ICON-CLM-simulated atmospheric variables using six machine learning (ML) models to detect lightning activity over the Third Pole. Results from the ensemble boosting ML models show that ICON-CLM simulated variables such as relative humidity (RH), vorticity (vor), 2m temperature (t_2m), and surface pressure (sfc_pres) among a total of 25 variables allow better spatial and temporal prediction of lightning activities, achieving a Probability of Detection (POD) of ∼0.65. The Lightning Potential Index (LPI) and the product of convective available potential energy (CAPE) and precipitation (prec_con), referred to as CP (i.e., CP = CAPE × precipitation), serve as key physics aware predictors, maintaining a high Probability of Detection (POD) of ∼0.62 with a 1–2 h lead time. Sensitivity analyses additionally using climatological lightning data showed that while ML models maintain comparable accuracy and POD, climatology primarily supports broad spatial patterns rather than fine-scale prediction improvements. As LPI and CP reflect cloud microphysics and atmospheric stability, their inclusion, along with spatiotemporal averaging and climatology, offers slightly lower, yet comparable, predictive skill to that achieved by aggregating 25 atmospheric predictors. Model evaluation using the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) highlights XGBoost as the best-performing diagnostic classification (yes/no lightning) model across all six ML tested configurations. Full article

Accurate and stable power-load forecasting is crucial for optimizing generation scheduling and ensuring the economic and secure operation of power grids. To address the issues of low prediction accuracy and poor robustness during abrupt load changes, this study proposes a Bayesian-optimized GCN-BiLSTM-Adaboost model (abbreviated as GCN-BiLSTM-AB). It combines Graph Convolutional Networks (GCN), Bidirectional Long Short-Term Memory Networks (BiLSTM), and a Bayesian-optimized AdaBoost framework. Firstly, the GCN is employed to capture the spatial correlation features of the input data. Then, the BiLSTM is employed to extract the long-term dependencies of the data time series. Finally, the AdaBoost framework is used to dynamically adjust the base learner weights, and a Bayesian method is employed to optimize the weight adjustment process and prevent overfitting. The experiment results on actual load data from a regional power grid show the GCN-BiLSTM-AB outperforms other compared models in prediction error metrics, with $MAE$, $MAPE$, and $RMSE$ values of 1.86, 3.13%, and 2.26, respectively, which improve the prediction robustness during load change periods. Therefore, the proposed method shows that the synergistic effect of spatiotemporal feature extraction and dynamic weight adjustment improves prediction accuracy and robustness, which provides a new forecasting model with high precision and reliability for power system dispatch decisions. Full article