Search Results (50)

Objective: To evaluate visual outcomes, rotational stability, patients’ satisfaction, and spectacle independence after bilateral Toric extended depth of focus intraocular lens (EDOF IOL) implantation. Methods: Prospective observational study including cataract patients with bilateral corneal astigmatism between 0.75 and 3.00 D implanted with Toric EDOF IOLs. After three months distance corrected and uncorrected visual acuity at 4 m (DCVA and UDVA), 80 cm (DCI80VA and UI80VA), 67 cm (DCI67VA and UI67VA), and 40 cm (DCNVA and UNVA), IOL stability by Toric IOL Assistant tool (Osiris T, CSO, Florence, Italy), binocular defocus curves, contrast sensitivity (CS), halometry, reading performance, and subjective and objective (Root mean square-RMS, modulation transfer function-MTF, cut-off and point-spread-function-PSF-Strehl ratio) visual quality were evaluated. Results: Forty eyes from 20 astigmatic patients were enrolled. Mean refractive spherical equivalent and residual cylinder were −0.21 ± 0.74 D and 0.29 ± 0.31 D, respectively. No patients needed additional surgery due to IOL rotation. Binocular UDVA, UI80VA, UI67VA, and UNVA ≤ 0.2 logMAR was found in 90%, 95%, 85%, and 80%. Distance-corrected visual outcomes have overall shown higher performances. All visual acuities at defocus curves were ≤0.125 logMAR between +0.50 D and −2.00 D. PSF-Strehl ratio, MTF cut-off, RMS were 0.26 ± 0.28, 19.82 ± 12.35, 0.31 ± 0.17. Reading analysis reached 125.42 ± 27.21 words/minute, 92.56 ± 7.82, 0.17 ± 0.15 logMAR and 0.50 ± 0.11 logRAD for mean reading speed, visual acuity score, reading acuity, and critical print size, respectively. CS was higher in photopic conditions. Subjective spectacle independence was achieved in 80% of patients. Conclusions: Toric EDOF IOL showed rotational stability and reliable astigmatic correction. It provided spectacle independence and good performance from distance to near distance, reaching high patient satisfaction without undermining binocular quality of vision. Full article

A significant challenge that hinders advancements in medical research is the sensitive and confidential nature of patient data in available datasets. In particular, sharing patients’ facial images poses considerable privacy risks, especially with the rise of generative artificial intelligence (AI), which could misuse such data if accessed by unauthorized parties. However, facial expressions are a valuable source of information for doctors and researchers, which creates a need for methods to derive them without compromising patient privacy or safety by exposing identifiable facial images. To address this, we present a quick, computationally efficient method for detecting action units (AUs) and their intensities—key indicators of health and emotion—using only 3D facial landmarks. Our proposed framework extracts 3D face landmarks from video recordings and employs a lightweight neural network (NN) to identify AUs and estimate AU intensities based on these landmarks. Our proposed method reaches a 79.25% F1-score in AU detection for the main AUs, and 0.66 in AU intensity estimation Root Mean Square Error (RMSE). This performance shows that it is possible for researchers to share 3D landmarks, which are far less intrusive, instead of facial images while maintaining high accuracy in AU detection. Moreover, to showcase the usefulness of our AU detection model, using the detected AUs and estimated intensities, we trained state-of-the-art Deep Learning (DL) models to detect pain. Our method reaches 91.16% accuracy in pain detection, which is not far behind the 93.14% accuracy obtained when employing a convolutional neural network (CNN) with residual blocks trained on actual images and the 92.11% accuracy obtained when employing all the ground-truth AUs. Full article

Over the last century, significant climate changes, including more intense droughts and floods, have impacted agriculture and socio-economic development, particularly in rain-dependent regions like Marrakech–Safi (MS) in Morocco. Limited data availability complicates the accurate monitoring and assessment of these natural hazards. This study evaluates the role of satellite data in drought monitoring in the MS region using rain gauge observations from 18 stations, satellite-based precipitation estimates from Climate Hazards Group InfraRed Precipitation with Station (CHIRPS), and temperatures from the fifth generation of the atmospheric global climate reanalyzed Era5-Land data. The Standardized Precipitation Evapotranspiration Index (SPEI) was calculated at various timescales to characterize droughts. Statistical analysis was then performed to assess the correlation between the SPEI and the cereal yields. The results show that CHIRPS effectively monitors droughts, demonstrating strong statistically significant correlations (r ~ 0.9) with the observed data in the plains, the plateaus, Essaouira–Chichaoua Basin, and the coastal zones, along with a good BIAS score and lower root mean square error (RMSE). However, discrepancies were observed in the High Atlas foothills and the mountainous regions. Correlation analysis indicates the significant impact of droughts on agricultural productivity, with strong correlations between the Standardized Yield Residual Series (SYRS) and SPEI-6 in April and SPEI-12 in June (r ~ 0.80). These findings underscore the importance of annual and late-season precipitation for cereal yields. Analysis provides valuable insights for decision-makers in designing adaptation strategies to enhance small-scale farmers’ resilience to current and projected droughts. Full article

Background/Objectives: Le Fort I maxillary osteotomy (LF1-MO) is associated with a risk of infraorbital nerve neurosensory deficiency (NSD). This study aimed to evaluate post-operative subjective numbness and objective NSD after LF1-MO and assess the impact of these outcomes on overall patient satisfaction. Methods: A retrospective cross-sectional study was conducted among adult LF1-MO patients, who were evaluated for treatment satisfaction using a 10-item patient satisfaction questionnaire. In addition, subjective and objective NSDs were assessed post-operatively for six months. Overall patient satisfaction was compared against different variables (patient age, sex, and type of LF1-MO) and NSD. The proportion of subjective and objective NSDs were statistically correlated and compared against these variables, assuming a 95% significance level (p < 0.05). Results: A total of 58 LF1-MO patients in the age range of 20–38 years (mean–29.79 ± 4.62 years) were included in this study. Most patients were females (n = 48; 82.8%) and aged 30 years and older (n = 32; 55.2%). The overall mean patient satisfaction score was 27.38 ± 3.94 (range 12–30), which did not significantly differ based on patient age or sex. Patients who had advanced LF1-MO had significantly higher satisfaction scores (28.27 ± 1.85) compared to those who had impaction (24.61 ± 7.34) (p < 0.05). Subjective numbness and an abnormal “Level A” response to objective neurosensory testing were associated with poor patient satisfaction. There was significant statistical correlation between subjective and objective NSDs (Spearman’s rho–0.441; p < 0.01). Based on a chi-squared test, patients undergoing maxillary setback (subjective–88.9%; objective–44.5%) had significantly higher NSDs (p < 0.05). Conclusions: Most patients reported satisfaction after LF1-MO, particularly among females, those aged 30 and older, and those without NSD. However, residual infraorbital NSDs persisted, with about two-thirds experiencing subjective numbness and 25% showing abnormal responses in “Level A” objective neurosensory tests six months post-operatively. Moreover, subjective numbness correlated with abnormal objective testing results, leading to lower patient satisfaction. Full article

This study investigates the operational efficiency of the lab-scale oxidation ditch (OD) functioning in simultaneous nitrification and denitrification modes, focusing on forecasting biochemical oxygen demand (BOD₅) concentrations over a five-day horizon. This forecasting capability aims to optimize the operational regime of aeration tanks by adjusting the specific load on organic pollutants through active sludge dosage modulation. A comprehensive statistical analysis was conducted to identify trends and seasonality alongside significant correlations between the forecasted values and various time lags. A total of 20 time lags and the “month” feature were selected as significant predictors. These models employed include Multi-head Attention Gated Recurrent Unit (MAGRU), long short-term memory (LSTM), Autoregressive Integrated Moving Average–Long Short-Term Memory (ARIMA–LSTM), and Prophet and gradient boosting models: CatBoost and XGBoost. Evaluation metrics (Mean Squared Error (MSE), Mean Absolute Error (MAE), Symmetric Mean Absolute Percentage Error (SMAPE), and Coefficient of Determination (R²)) indicated similar performance across models, with ARIMA–LSTM yielding the best results. This architecture effectively captures short-term trends associated with the variability of incoming wastewater. The SMAPE score of 1.052% on test data demonstrates the model’s accuracy and highlights the potential of integrating artificial neural networks (ANN) and machine learning (ML) with mechanistic models for optimizing wastewater treatment processes. However, residual analysis revealed systematic overestimation, necessitating further exploration of significant predictors across various datasets to enhance forecasting quality. Full article

Radar echo extrapolation by ground-based remote sensing is essential for weather prediction and flight guiding. Existing radar echo extrapolation methods can hardly capture complex spatiotemporal features, resulting in the low accuracy of predictions, and, therefore, severely restrict their use in extreme weather situations. A deep learning method was recently applied for extrapolating radar echoes; however, its accuracy declines too quickly over a short time. In this study, we introduce a solution: Residual Transformer and Unet (ResTUnet), a novel model that improves prediction accuracy and exhibits good stability with a slow rate of accuracy decline. This presented Rest-Net model is designed to solve the issue of declining prediction accuracy by integrating a 1*1 convolution to diminish the neural network parameters. We constructed an observed dataset by Zhengzhou East Airport radar observation from July 2022 to August 2022 and performed 90 min experiments comprising five aspects, including extrapolation images, the Probability of Detection (POD) index, the Critical Success Index (CSI), the False Alarm Rate (FAR) index, and the Heidke Skill Score (HSS) index. The experimental results show that the ResTUnet model improved the CSI, HSS index, and the POD index by 17.20%, 11.97%, and 11.35%, compared to current models, including Convolutional Long Short-Term Memory (convLSTM), the Convolutional Gated Recurrent Unit (convGRU), the Trajectory Gated Recurrent Unit (TrajGRU), and the improved recurrent network for video predictive learning, the Predictive Recurrent Neural Network++ (predRNN++). In addition, the mean squared error of the ResTUnet model remains stable at 15% between 0 and 60 min and starts to increase after 60–90 min, which is 12% better than the current models. This enhancement in prediction accuracy has practical applications in meteorological services and decision making. Full article

Predicting nutrient loads is essential to understanding and managing one of the environmental issues faced by the northern Gulf of Mexico hypoxic zone, which poses a severe threat to the Gulf’s healthy ecosystem and economy. The development of hypoxia in the Gulf of Mexico is strongly associated with the eutrophication process initiated by excessive nutrient loads. Due to the complexities in the excessive nutrient loads to the Gulf of Mexico, it is challenging to understand and predict the underlying temporal variation of nutrient loads. The study was aimed at identifying an optimal predictive machine learning model to capture and predict nonlinear behavior of the nutrient loads delivered from the Mississippi/Atchafalaya River Basin (MARB) to the Gulf of Mexico. For this purpose, monthly nutrient loads (N and P) in tons were collected from US Geological Survey (USGS) monitoring station 07373420 from 1980 to 2020. Machine learning models—including autoregressive integrated moving average (ARIMA), gaussian process regression (GPR), single-layer multilayer perceptron (MLP), and a long short-term memory (LSTM) with the single hidden layer—were developed to predict the monthly nutrient loads, and model performances were evaluated by standard assessment metrics—Root Mean Square Error (RMSE) and Correlation Coefficient (R). The residuals of predictive models were examined by the Durbin–Watson statistic. The results showed that MLP and LSTM persistently achieved better accuracy in predicting monthly TN and TP loads compared to GPR and ARIMA. In addition, GPR models achieved slightly better test RMSE score than ARIMA models while their correlation coefficients are much lower than ARIMA models. Moreover, MLP performed slightly better than LSTM in predicting monthly TP loads while LSTM slightly outperformed for TN loads. Furthermore, it was found that the optimizer and number of inputs didn’t show effects on the LSTM performance while they exhibited impacts on MLP outcomes. This study explores the capability of machine learning models to accurately predict nonlinearly fluctuating nutrient loads delivered to the Gulf of Mexico. Further efforts focus on improving the accuracy of forecasting using hybrid models which combine several machine learning models with superior predictive performance for nutrient fluxes throughout the MARB. Full article

The identification of significant gene biclusters with particular expression patterns and the elucidation of functionally related genes within gene expression data has become a critical concern due to the vast amount of gene expression data generated by RNA sequencing technology. In this paper, a Conserved Gene Expression Module based on Genetic Algorithm (CGEMGA) is proposed. Breast cancer data from the TCGA database is used as the subject of this study. The p-values from Fisher’s exact test are used as evaluation metrics to demonstrate the significance of different algorithms, including the Cheng and Church algorithm, CGEM algorithm, etc. In addition, the F-test is used to investigate the difference between our method and the CGEM algorithm. The computational cost of the different algorithms is further investigated by calculating the running time of each algorithm. Finally, the established driver genes and cancer-related pathways are used to validate the process. The results of 10 independent runs demonstrate that CGEMGA has a superior average p-value of 1.54 × 10⁻⁴ ± 3.06 × 10⁻⁵ compared to all other algorithms. Furthermore, our approach exhibits consistent performance across all methods. The F-test yields a p-value of 0.039, indicating a significant difference between our approach and the CGEM. Computational cost statistics also demonstrate that our approach has a significantly shorter average runtime of 5.22 × 10⁰ ± 1.65 × 10⁻¹ s compared to the other algorithms. Enrichment analysis indicates that the genes in our approach are significantly enriched for driver genes. Our algorithm is fast and robust, efficiently extracting co-expressed genes and associated co-expression condition biclusters from RNA-seq data. Full article

In the ongoing battle against antibiotic-resistant bacteria, New Delhi metallo-β-lactamase-1 (NDM-1) has emerged as a significant therapeutic challenge due to its ability to confer resistance to a broad range of β-lactam antibiotics. This study presents a pharmacophore-based virtual screening, docking, and molecular dynamics simulation approach for the identification of potential inhibitors targeting NDM-1, a critical enzyme associated with antibiotic resistance. Through the generation of a pharmacophore model and subsequent virtual screening of compound libraries, candidate molecules (ZINC29142850 (Z1), ZINC78607001 (Z2), and ZINC94303138 (Z3)) were prioritized based on their similarity to known NDM-1 binder (hydrolyzed oxacillin (0WO)). Molecular docking studies further elucidated the binding modes and affinities of the selected compounds towards the active site of NDM-1. These compounds demonstrated superior binding affinities to the enzyme compared to a control compound (−7.30 kcal/mol), with binding scores of −7.13, −7.92, and −8.10 kcal/mol, respectively. Binding interactions within NDM-1’s active site showed significant interactions with critical residues such as His250, Asn220, and Trp93 for these compounds. Subsequent molecular dynamics simulations were conducted to assess the stability of the ligand–enzyme complexes, showing low root mean square deviation (RMSD) values between 0.5 and 0.7 nm for Z1, Z2, which indicate high stability. Z2’s compactness in principal component analysis (PCA) suggests that it can stabilize particular protein conformations more efficiently. Z2 displays a very cohesive landscape with a notable deep basin, suggesting a very persistent conformational state induced by the ligand, indicating robust binding and perhaps efficient inhibition. Z2 demonstrates the highest binding affinity among the examined compounds with a binding free energy of −25.68 kcal/mol, suggesting that it could offer effective inhibition of NDM-1. This study highlights the efficacy of computational tools in identifying novel antimicrobial agents against resistant bacteria, accelerating drug discovery processes. Full article

This study examines the performance of various machine learning (ML) models in predicting Interstitial Glucose (IG) levels using data from wrist-worn wearable sensors. The insights from these predictions can aid in understanding metabolic syndromes and disease states. A public dataset comprising information from the Empatica E4 smart watch, the Dexcom Continuous Glucose Monitor (CGM) measuring IG, and a food log was utilized. The raw data were processed into features, which were then used to train different ML models. This study evaluates the performance of decision tree (DT), support vector machine (SVM), Random Forest (RF), Linear Discriminant Analysis (LDA), K-Nearest Neighbors (KNN), Gaussian Naïve Bayes (GNB), lasso cross-validation (LassoCV), Ridge, Elastic Net, and XGBoost models. For classification, IG labels were categorized into high, standard, and low, and the performance of the ML models was assessed using accuracy (40–78%), precision (41–78%), recall (39–77%), F1-score (0.31–0.77), and receiver operating characteristic (ROC) curves. Regression models predicting IG values were evaluated based on R-squared values (?7.84–0.84), mean absolute error (5.54–60.84 mg/dL), root mean square error (9.04–68.07 mg/dL), and visual methods like residual and QQ plots. To assess whether the differences between models were statistically significant, the Friedman test was carried out and was interpreted using the Nemenyi post hoc test. Tree-based models, particularly RF and DT, demonstrated superior accuracy for classification tasks in comparison to other models. For regression, the RF model achieved the lowest RMSE of 9.04 mg/dL with an R-squared value of 0.84, while the GNB model performed the worst, with an RMSE of 68.07 mg/dL. A SHAP analysis identified time from midnight as the most significant predictor. Partial dependence plots revealed complex feature interactions in the RF model, contrasting with the simpler interactions captured by LDA. Full article

Creating an effective deep learning technique for accurately diagnosing leak signals across diverse environments is crucial for integrating artificial intelligence (AI) into the power plant industry. We propose an automatic weight redistribution ensemble model based on transfer learning (TL) for detecting leaks in diverse power plant environments, overcoming the challenges of site-specific AI methods. This innovative model processes time series acoustic data collected from multiple homogeneous sensors located at different positions into three-dimensional root-mean-square (RMS) and frequency volume features, enabling accurate leak detection. Utilizing a TL-driven, two-stage learning process, we first train residual-network-based models for each domain using these preprocessed features. Subsequently, these models are retrained in an ensemble for comprehensive leak detection across domains, with control weight ratios finely adjusted through a softmax score-based approach. The experiment results demonstrate that the proposed method effectively distinguishes low-level leaks and noise compared to existing techniques, even when the data available for model training are very limited. Full article

Eating outside-of-home (EOH) is one of the main changes in lifestyle that occurred worldwide in the past few decades. Given that EOH behavior is influenced by individual and contextual factors, the utilization of a theory seems to be suitable in analyzing this health behavior. The fourth-generation theory multi-theory model (MTM) is designed exclusively for health behavior change at the individual and community levels. Therefore, the purpose of this analytical cross-sectional study was to investigate EOH behavior by using the MTM among a nationally representative sample in the United States (US). Data for this study were collected from April–May 2023 via a 61-item psychometric valid, web-based, structured survey disseminated via Qualtrics. Chi-square/Fisher’s exact tests were used to compare categorical data, whereas the independent-samples t-test was used to compare the mean scores of MTM constructs across groups. Pearson correlation analysis was performed for the intercorrelation matrix between the MTM constructs and hierarchical regression models were built to predict the variance in the initiation and sustenance by certain predictor variables beyond demographic characteristics. The p values in the multiple comparisons were calculated by using adjusted residuals. Among a total of 532 survey respondents, 397 (74.6%) indicated being engaged in EOH at least twice a week, whereas 135 (25.4%) reported not being engaged in EOH. People who were engaged in EOH were younger (mean age = 42.25 ± 17.78 years vs. 55.89 ± 19.43 years) African American, (15.9% vs. 6.7%, p = 0.01), single or never married, (34.0% vs. 23.0%, p = 0.02), had a graduate degree (9.6% vs. 3.7%, p = 0.03), and were employed (72.0% vs. 34.8%, p < 0.001) as opposed to those who reported not being engaged in eating outside the home. Among the MTM constructs of initiation, “behavioral confidence” and “changes in the physical environment” were the significant predictors of initiating a reduction in EOH behavior and explained 48% of the variance in initiation. Among the MTM constructs of sustenance, “emotional transformation” and “changes in the social environment” were the significant predictors of sustaining a reduction in EOH behavior and explained 50% of the variance in sustenance. This study highlights a need to design MTM-based educational interventions that promote in-home eating instead of frequent EOH for health, family bonding, economic, and other reasons. Full article

A tropical cyclone (TC) is a type of severe weather system that damages human property. Understanding TC mechanics is crucial for disaster management. In this study, we propose a multi-task learning framework named Multi-Task Graph Residual Network (MT-GN) to classify and estimate the intensity of TCs from FY-4A geostationary meteorological satellite images. And we construct a new benchmark dataset collected from the FY-4A satellite for both TC classification and intensity estimation tasks. Four different methodologies to classify TCs and estimate the intensity of TCs are fairly compared in our dataset. We discover that accurate classification and estimation of TCs, which are usually achieved separately, requires co-related knowledge from each process. Thus, we train a convolution feature extractor in a multi-task way. Furthermore, we build a task-dependency embedding module using a Graph Convolution Network (GCN) that further drives our model to reach better performance. Finally, to overcome the influence of the unbalanced distribution of TC category samples, we introduce class-balanced loss to our model. Experimental results on the dataset show that the classification and estimation performance are improved. With an overall root mean square error (RMSE) of 9.50 knots and F1-score of 0.64, our MT-GN model achieves satisfactory performance. The results demonstrate the potential of applying multi-task learning for the study of TCs. Full article

Aiming at the problem of the remaining useful life prediction accuracy being too low due to the complex operating conditions of the aviation turbofan engine data set and the original noise of the sensor, a residual useful life prediction method based on spatial–temporal similarity calculation is proposed. The first stage is adaptive sequence matching, which uses the constructed spatial–temporal trajectory sequence to match the sequence to find the optimal matching sample and calculate the similarity between the two spatial–temporal trajectory sequences. In the second stage, the weights of each part are assigned by the two weight allocation algorithms of the weight training module, and then the final similarity is calculated by the similarity calculation formula of the life prediction module, and the final predicted remaining useful life is determined according to the size of the similarity and the corresponding remaining life. Compared with a single model, the proposed method emphasizes the consistency of the test set and the training set, increases the similarity between samples by sequence matching with other spatial–temporal trajectories, and further calculates the final similarity and predicts the remaining use through the weight allocation module and the life prediction module. The experimental results show that compared with other methods, the root mean square error (RMSE) index and the remaining useful life health score (Score) index are reduced by 12.6% and 14.8%, respectively, on the FD004 dataset, and the RMSE index is similar to that in other datasets; the Score index is reduced by about 10%, which improves the prediction accuracy of the remaining useful life and can provide favorable support for the operation and maintenance decision of turbofan engines. Full article

Zearalenone (ZEN) is one of the most prevalent estrogenic mycotoxins, is produced mainly by the Fusarium family of fungi, and poses a risk to the health of animals. Zearalenone hydrolase (ZHD) is an important enzyme capable of degrading ZEN into a non-toxic compound. Although previous research has investigated the catalytic mechanism of ZHD, information on its dynamic interaction with ZEN remains unknown. This study aimed to develop a pipeline for identifying the allosteric pathway of ZHD. Using an identity analysis, we identified hub genes whose sequences can generalize a set of sequences in a protein family. We then utilized a neural relational inference (NRI) model to identify the allosteric pathway of the protein throughout the entire molecular dynamics simulation. The production run lasted 1 microsecond, and we analyzed residues 139–222 for the allosteric pathway using the NRI model. We found that the cap domain of the protein opened up during catalysis, resembling a hemostatic tape. We used umbrella sampling to simulate the dynamic docking phase of the ligand–protein complex and found that the protein took on a square sandwich shape. Our energy analysis, using both molecular mechanics/Poisson–Boltzmann (Generalized-Born) surface area (MMPBSA) and Potential Mean Force (PMF) analysis, showed discrepancies, with scores of −8.45 kcal/mol and −1.95 kcal/mol, respectively. MMPBSA, however, obtained a similar score to that of a previous report. Full article