Search Results (4,396)

(1) Background: One-lung ventilation (OLV) is frequently required during thoracic surgery, but hypoxemia remains a common intraoperative complication. Neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) have emerged as inexpensive inflammatory biomarkers, although their role in predicting hypoxemia during OLV remains unclear. This study evaluated the association between preoperative NLR, PLR, and severe intraoperative hypoxemia during OLV. (2) This interim analysis included 103 patients undergoing elective thoracic surgery with OLV in a prospective observational cohort. Severe hypoxemia was defined as PaO₂/FiO₂ < 100. Group comparisons were performed using Mann–Whitney U and chi-square/Fisher’s exact tests. Hierarchical logistic regression and ROC analysis were used to evaluate predictors and model performance. (3) Results: Preoperative PLR significantly improved the predictive performance of the clinical model for severe intraoperative hypoxemia, while NLR was not associated with the outcome. BMI remained an independent predictor of hypoxemia. (4) Conclusions: PLR improved the predictive performance of the clinical model, although its inverse association with hypoxemia should be interpreted cautiously. NLR was not associated with hypoxemia during OLV. Full article

In the past few years, cyber-attacks have risen at an exponential rate across all sectors, and both private and public institutions have faced increasingly sophisticated threats. As this upward trend continues, the need for advanced and efficient threat detection systems is essential. This paper investigates the use of feature importance (FI) Coefficients to improve Artificial Neural Network (ANN) and Convolutional Neural Network (CNN) models, leveraging feature selection to enhance model interpretability and optimize performance. By systematically filtering out the weaker features, we examine the reduced features’ impact on model accuracy, precision, recall, and F1 score. Experiments were conducted on two new datasets, UWF-ZeekDataSum2025-1 and UWF-ZeekDataSum2025-2, using a baseline ANN/CNN architecture and multiple architectural variants. The results on UWF-ZeekDataSum2025-1 show a clear performance gain for certain feature importance thresholds, with models such as ANN-Minimal, ANN-Overfit-Wide, ANN-Shallow-Low-Optimization, CNN-Shallow, and CNN-Very-Shallow outperforming the baseline after reducing the feature space from seventeen features to fewer than four. For UWF-ZeekDataSum2025-2, improvements occur across a broader range of thresholds, with models including ANN-Deep-Sub-Conv, ANN-Shallow-Low-Opt, CNN-Shallow, CNN-Very-Shallow, and ANN-Minimal exceeding 95% performance around the 0.25–0.28 thresholds, with additional gains at 0.31–0.32 for some architectures. These findings demonstrate that by strategically leveraging feature importance coefficient thresholds, we can significantly enhance neural network intrusion detection systems, offering a reproducible pathway for adapting these methods on similar environments. Full article

The prickly pear is a crop of socioeconomic relevance in arid regions, and its productivity and chemical composition depend on water availability. The effect of irrigation on the crop’s biochemical quality was evaluated. Cladodes of cultivars: ‘Amarilla Olorosa’, ‘Cristalina’, ‘Dalia Roja’, and ‘Roja Lisa’, were subjected to three treatments: no irrigation (NI), supplemental irrigation (SI), equivalent to 50% of the crop’s evapotranspiration, and full irrigation (FI). Subsequently, cladodes were collected, and total polyphenols and flavonoids, polyphenol profile, and antioxidant capacity were determined. Cladodes under NI had the highest concentrations of flavonoids, although the lowest values of total polyphenols. In the cladode extracts, myricetin, rutin, catechin, as well as caffeic, chlorogenic, dihydroxybenzoic, and vanillic acids were identified. Overall, cladodes grown under FI and SI showed higher levels of phenolic acids (caffeic, chlorogenic, and vanillic), while concentrations of catechin, myricetin, and rutin were higher under SI and NI. Antioxidant capacity was higher in NI cladodes assessed by ABTS and DPPH, while the FRAP assay showed higher values under SI. Among the cultivars, ‘Amarilla Olorosa’ stood out for its high content of bioactive compounds, confirming the potential of nopal cladodes as a source of antioxidant metabolites with agro-industrial applications. Full article

Horticultural intensification in Mediterranean areas has increased the risk of nitrate pollution due to inefficient irrigation and nitrogen fertilisation management. The availability of simulation models aimed at rational nitrogen management in outdoor crops is limited. The objective of this study is to adapt the VegSyst model, initially developed for greenhouse vegetables, for use in open-field conditions in relevant crops, such as processing peppers and broccoli in Extremadura. VegSyst simulates dry matter production and nitrogen uptake by incorporating the influence of evaporative demand (TUE approach) in addition to the effect of radiation (RUE approach). Experimental field data obtained in five campaigns (peppers: 2020–2022; broccoli: 2020 and 2022) under different nitrogen doses were used. The model was calibrated, and critical N dilution curves were developed for each crop. Subsequently, the simulation of fi-PAR, dry matter production (DMP) and N uptake was validated using statistical indices (RMSE, RE, d, EF) and regression analysis. The model showed a high predictive capacity for N uptake in both crops, with values of d ≥ 0.98 and EF ≥ 0.90 in the validation campaigns. The fi-PAR simulation was acceptable in peppers and excellent in broccoli. In contrast, the DMP prediction showed notable deviations in peppers, especially in 2022, attributable to interannual variations in weather conditions and physiological limitations not considered by the model. In both crops, the TUE-based strategy was a better fit for the measurements than the RUE-based strategy, indicating that under semi-arid Mediterranean conditions, transpiration is the limiting factor for biomass production. The adaptation of the VegSyst-Outdoors model proved to be robust for simulating N uptake and sufficiently accurate to be integrated into decision support tools aimed at efficient fertilisation and irrigation management. Full article

Accurate real-time soil moisture monitoring is critical for optimizing water use and ensuring crop health and food security. This study aims to calibrate and integrate low-cost capacitive soil moisture sensors into a solar-powered sensor node for real-time soil moisture monitoring in a loamy sand soil. Three capacitive soil moisture sensors were calibrated in the laboratory under controlled volumetric water content conditions (0–40%) using a constrained linear regression approach. The system was tested in a limited pilot-scale in a drip-irrigated onion field at the IWAD farm, Yagaba (North-East Region, Ghana). The results showed good agreement of the sensor readings with the soil moisture obtained using the gravimetric method (R² of 0.92–0.94, RMSE of 0.40–0.52%, and MAE of 0.35–0.39%) demonstrating the successful transfer of the calibration functions to field conditions. Soil moisture data was successfully monitored and transmitted from the nodes to a LoRa gateway via LoRaWAN (433 MHz) and from the gateway to a Raspberry Pi edge server via Wi-Fi. Data was stored both locally in SQLite on the Raspberry Pi and on the InfluxDB cloud. These results suggest that the developed system, when extensively validated under field conditions, can be used to support decision-making for data-driven IoT-based irrigation scheduling. Full article

Background/Objectives: We investigated the association of serum ferritin and procalcitonin (PCT) with Sepsis-related Organ Failure Assessment (SOFA)-2 score-based organ dysfunction severity in intensive care patients with COVID-19-associated sepsis. Methods: Patients were stratified by day 5 ferritin (ng/mL) and PCT (μg/L) levels; associations were analysed across severity groups defined by an SOFA-2 score of <5 (mild) or ≥5 (severe). Results: Day 5 PCT did not predict the SOFA-2 score (p > 0.05). The optimal day 5 ferritin cut-off was >1191 ng/mL (35.78% sensitivity, 82.38% specificity; area under the curve (AUC) = 0.608). Day 5 ferritin was associated with SOFA-2 severity in the univariable analysis but did not remain an independent correlate after adjustment for C-reactive protein (CRP) and lactate dehydrogenase (LDH); in a mortality model, neither ferritin nor PCT independently predicted intensive care unit (ICU) death. PCT provided no predictive value beyond existing inflammatory markers, consistent with its suppression during viral infections. Conclusions: Day 5 ferritin reflects, rather than independently predicts, organ dysfunction severity and may complement, rather than replace, established multi-parameter scoring. Relative to the independent determinants of severity and mortality (PaO₂/FiO₂ ratio, LDH, CRP, and age), day 5 ferritin is a specific, rule-in adjunctive marker of concurrent organ dysfunction rather than a standalone prognostic tool. Whether these associations extend to non-COVID sepsis populations requires prospective study. Full article

This study investigates whether financial sector development, and, critically, which dimension of it, is associated with the dual energy transition across eleven Southeast Asian economies over 2004–2020. The empirical strategy combines Pooled OLS with Driscoll–Kraay standard errors, two-way fixed effects, Pooled Mean Group ARDL error correction, and Method-of-Moments quantile regression. The results reveal a stark asymmetry: the Financial Institutions Index is positively and robustly associated with clean cooking access across all estimators. Quantile regressions confirm that the FI association with clean cooking is significant across the entire distribution, with the largest coefficients at the lower quantiles. Sub-sample analysis reveals that the FI–clean cooking relationship is especially pronounced in the frontier Cambodia–Lao PDR–Myanmar–Vietnam–Timor-Leste group, where within-country fixed effects yield a coefficient of 257.54 (p < 0.01). Although these associations do not establish strict causality, the findings are consistent with prioritising deepening institutional banking and digital financial inclusion rather than equity-market development as the primary financial-sector channel associated with lower energy poverty in Southeast Asia, although such policy directions require further micro-level validation. Full article

Introduction: Restoring riverine connectivity is a cornerstone of ecological restoration for migratory fish populations. When physical barriers like dams lack effective fishways, translocation to more suitable sites becomes an alternative. Objectives: This study aims to present a decision-support methodology based on the Fragmentation Index (FI), designed to prioritize release sites in alternative river stretches that maximize the likelihood of survival of translocated diadromous fish. Methodology: The method integrates field-based obstacle characterization and transposability classification, together with a weighted penalty for restrictive obstacles located closer to the confluence with the main stem. The methodology was applied to six tributaries of the Douro River, targeting the European eel (Anguilla anguilla). Results: The FI successfully distinguished between functional reaches and severely fragmented systems. Results revealed high heterogeneity among the studied tributaries, with the Távora (FI = 1.07) and Ceira (FI = 1.12) Rivers identified as top priorities due to low fragmentation and stable hydrology. In contrast, the Tedo River (FI = 5.18) illustrates index’s sensitivity. Despite a high barrier density, its downstream stretch of ~14 km remains functionally connected because the first restrictive obstacles are located far upstream from the confluence. Conversely, the Torto River (FI = 0) was excluded due to severe drought conditions, underscoring the need to pair connectivity metrics with hydrologic viability. Conclusions: For large-scale translocations, this framework enables distributing fish across multiple systems to safeguard the ecological integrity of recipient communities while ensuring individuals can successfully complete their life cycles. Overall, this approach provides a quantitative and replicable framework for managing endangered species by prioritizing release sites with high longitudinal connectivity. Full article

Smart energy meters are widely deployed in modern distribution networks, extending their role beyond revenue billing to real-time monitoring and data-driven smart city applications. However, conventional legal metrology frameworks rely on periodic recalibration and are not intended for the detection of accuracy drift or unexpected malfunctions between scheduled inspections. In scientific publications, various techniques for remote smart meters’ error surveillance are presented, but experimental verification on real distribution network data remains limited. The objective of this study is to experimentally verify two previously proposed power event-driven methods for remote estimation of active power measurement error in individual consumer meters, using a feeder-level sum meter as a reference instrument. One-second resolution electrical readings were collected from a real low-voltage distribution branch using ESP32-based local adapters communicating via MQTT over Wi-Fi, with SNTP-based clock synchronization for power event correlation. Under optimized detection parameters, the linear regression method achieved 0.20% RMSE and 0.75% maximum absolute error, and the neural network method 0.09% RMSE and 0.31%, confirming suitability for Class 1 m accuracy surveillance. Feasibility analysis of three MQTT-based deployment scenarios demonstrates that binary encoding limits local adapter buffers to 2.8 kB and worst-case daily channel demand to 2000 kB, confirming the practical viability of the proposed architecture. Full article

The digital transformation of tourist facilities requires careful selection of technologies that can provide secure, stable and scalable network infrastructure. Due to the possibility of application in different sectors with different specificities, the focus of the research was placed on the implementation of smart tourist services. A hybrid multi-criteria decision-making model based on PIPRECIA and MVA models was applied for the research. Based on the literature and the opinions of experts in the field, evaluation criteria such as bandwidth, latency, energy efficiency, security and privacy, scalability, costs and interoperability were defined, and internet technologies such as Li-Fi, Wi-Fi 7, Wi-Fi 6, private 5G networks, Ethernet-over-Power (EoP), NB-IoT and LoRaWAN were defined. The results obtained put the security and privacy criterion at the top (0.2253), followed by scalability (0.1952) and bandwidth (0.1624). The obtained results indicate that Wi-Fi 7 achieved the highest weighted score (4.2247), followed closely by Li-Fi (4.2177) and Wi-Fi 6 (4.0771). Wi-Fi 7 demonstrated particularly strong performance in scalability, interoperability and bandwidth, making it highly suitable for environments with high user density. Li-Fi achieved very high scores in security and latency, which makes it particularly appropriate for security-sensitive smart tourism environments. Lower-ranked technologies such as NB-IoT and LoRaWAN proved valuable for supporting IoT and monitoring functions, rather than as primary communication infrastructure. The proposed model has proven to be a flexible, transparent and practical tool for strategic decision-making in the field of smart tourism. In addition to the basic application presented in the paper, the model has the potential to be adapted to different contexts and expanded with additional criteria or new technologies. The proposed hybrid approach can serve as a useful decision-making tool for tourism managers, system engineers and urban planners who are looking for optimal solutions for the development of digital infrastructure. Full article

Background/Objectives: While packed red blood cell transfusions are commonly administered in anemic neonates, transfusion strategies in preterm infants have been the subject of debate for decades, particularly due to questionable long-term benefits and limited evidence regarding short-term physiological effects. In non-intubated preterm infants, established transfusion thresholds are considered, but individual clinical judgment often plays an important role in the final decision. This study aims to assess the short-term cardiorespiratory effects of red blood cell transfusions in non-intubated very-low-birth-weight (VLBW) infants who were either spontaneously breathing or receiving non-invasive respiratory support. Methods: Retrospective, single-center analysis of 68 VLBW infants (<1500 g) who received 99 red blood cell transfusions between 2019 and 2023. Cardiorespiratory parameters were observed over a 24 h period before and after transfusion. Results: Following transfusion, there was a significant decrease in the frequency of bradycardia events per 24 h (6.51 ± 5.55 to 4.24 ± 3.8; p = 0.004), accompanied by an improvement in the depth of oxygen desaturations (78.7 ± 4.18 to 81.0 ± 3.71; p = 0.001). No significant changes were detected in the desaturation frequency, FiO₂ or heart rate. Conclusions: In clinically stable very-low-birth-weight infants receiving non-invasive ventilatory support, packed red blood cell transfusion is associated with modest, short-term improvements in cardiorespiratory stability. However, these effects are limited in scope. Further research is needed to identify which patient subgroups derive the most significant benefit from these transfusions. Full article

In cold arid regions, the relationships between dissolved organic matter (DOM) characteristics and heavy metal distributions across ice, water, and sediment interfaces remain insufficiently resolved. This study characterized DOM spectral features and examined their associations with measured metal distributions in a typical frozen eutrophic lake using excitation–emission matrices coupled with parallel factor analysis (EEMs-PARAFAC), ultraviolet-visible absorption spectroscopy (UV-Vis), and Fourier-transform infrared spectroscopy (FTIR). Protein-like substances dominated ice DOM, whereas water and sediment-derived DOM contained more humified fluorescent components. Fluorescence indices confirmed a primarily biological origin across all media, with ice showing the highest autochthonous microbial contribution (BIX = 1.23) but the lowest humification (HIX = 0.26), suggesting a greater contribution of recently produced protein-like fluorescent DOM in the ice samples. Water DOM showed the highest average HIX (1.88), followed by sediment-derived DOM (0.61) and ice DOM (0.26). The measured hydrochemical conditions, including weak alkalinity, elevated total dissolved solids (TDS), and locally low dissolved oxygen, provide environmental context for differences in metal distributions. Exploratory Spearman analysis at 17 matched water stations identified the strongest DOM–metal associations for HIX-As (rho = 0.474, p = 0.054) and FI-Zn (rho = 0.471, p = 0.056), indicating that DOM optical properties provide testable indicators of metal-distribution patterns but should be combined with direct binding and speciation measurements for mechanistic confirmation. Because ice was collected in January 2021, whereas water and sediment were collected in October 2020, cross-medium differences are interpreted as between-campaign associations rather than synchronous partitioning. These findings provide a basis for targeted winter monitoring and future binding, speciation, and freeze-concentration experiments in shallow eutrophic lakes. Full article

Background: Frailty is associated with adverse postoperative outcomes and functional decline in older adults undergoing colorectal cancer (CRC) surgery. The long-term course of frailty and functional outcomes among patients undergoing prehabilitation before CRC surgery remains insufficiently investigated. Methods: This prospective observational cohort study evaluated long-term functional and physiological outcomes in older adults with frailty syndrome undergoing colorectal cancer (CRC) surgery who participated in a structured prehabilitation program. Forty-one patients aged >70 years were assessed before prehabilitation and at one-year follow-up. Frailty (the Clinical Frailty Scale [CFS] and the 5-item Frailty Index [5-FI]), physical activity, postural function, respiratory parameters, and functional performance (the 6 min walk test [6MWT] and the Timed Up and Go [TUG] test) were evaluated. Results: Of the 93 eligible patients, 41 completed the one-year follow-up and were therefore included in the final analysis. A small but statistically significant increase in frailty was observed using 5-FI (mean difference = 0.029, p = 0.012), with no significant change in CFS. Postural function improved (p = 0.031), while physical activity and functional performance remained stable (6MWT: 392.71 vs. 384.36 m, p = 0.885; TUG: 12.36 vs. 10.42 s, p = 0.051). A significant reduction in pre- and post-exercise oxygen saturation was observed; however, the magnitude of change (before: −1.25%, p = 0.006; after: −0.91%, p < 0.001) was small and of uncertain relevance. Conclusions: Over a one-year follow-up of prehabilitated CRC patients with frailty, their functional performance remained stable despite a subtle progression of frailty. These findings suggest a dissociation between physiological vulnerability and functional status. Due to the observational design of the study and the lack of a control group, the results should be interpreted as descriptive rather than causal. Full article

Background/Objectives: Postoperative hypoxemia is a frequent complication after non-cardiac surgery and is correlated with elevated mortality rates in elderly patients. However, a dedicated predictive tool for mortality in this specific patient subgroup remains unavailable. To construct and validate a machine learning (ML) model for predicting in-hospital mortality among elderly adults who develop hypoxemia after non-cardiac surgery. Methods: Data for this retrospective cohort study were obtained from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. The study encompassed patients aged 65 years or older who exhibited hypoxemia, defined as a PaO₂/FiO₂ ratio below 300 mmHg, within the initial 48 h of intensive care unit (ICU) stay. LASSO (Least Absolute Shrinkage and Selection Operator) regression was applied for feature selection, after which six distinct machine learning models and five conventional scoring systems were constructed and evaluated. SHapley Additive exPlanations (SHAP) was employed to improve model interpretability. Results: Out of 6051 eligible patients, 1838 (30.4%) succumbed during hospitalization. The XGBoost algorithm demonstrated superior predictive capability, achieving an area under the curve (AUC) of 0.794, along with a specificity of 0.917, accuracy of 0.769, and positive predictive value of 0.693. Critical predictors identified included administration of vasopressors, advanced age, and the PaO₂/FiO₂ ratio. Conclusions: The Extreme Gradient Boosting (XGBoost)-driven ML model provides accurate prediction of in-hospital mortality in elderly patients with postoperative hypoxemia after non-cardiac surgery, presenting a valuable instrument for early risk evaluation and potential intervention. Full article

Background: Cardiovascular disease and mortality are common outcomes of cardiovascular–kidney–metabolic (CKM) syndrome. The integrated role of metabolic dysfunction and frailty, quantified by the triglyceride–glucose–frailty index (TyG-FI), remains insufficiently explored. This study examined the association between TyG-FI and incident composite outcomes among participants with CKM stages 0–3. Methods: Data were obtained from two large cohort studies conducted in China and the United States. The analysis focused on participants classified as CKM stages 0–3. Cox proportional hazards models were used to estimate the relationship between TyG-FI and incident composite outcomes. Nonlinear associations were explored using spline functions. Additional analyses were performed across different subgroups and under varied assumptions. Model performance over time was also assessed. Results: Significant differences in outcome incidence were observed across TyG-FI levels. Higher quartiles showed a gradual increase in risk and displayed a dose–response pattern, with inflection points at 1.01 and 2.29. Associations were consistent across subgroups, and TyG-FI demonstrated moderate discrimination (AUCs 0.714 and 0.744). Conclusions: In the CHARLS and HRS cohorts, higher TyG-FI scores were independently associated with an increased risk of incident composite outcomes among participants with CKM stages 0–3, with a nonlinear relationship observed. Its discriminatory power was moderate, suggesting that TyG-FI may serve as a supplementary indicator for risk stratification in the early to mid-stages, although its clinical predictive value requires further validation. Full article