Search Results (4,208)

Background/Objectives: Non-variceal upper gastrointestinal bleeding (NVUGIB) is associated with considerable morbidity and mortality, particularly in emergency department (ED) settings. While traditional clinical scores such as the Glasgow-Blatchford Score (GBS), AIMS65, and Pre-Endoscopic Rockall are widely used for risk stratification, their accuracy in mortality prediction is limited. This study aimed to evaluate the performance of multiple supervised machine learning (ML) models in predicting 30-day all-cause mortality in NVUGIB and to compare these models with established risk scores. Methods: A retrospective cohort study was conducted on 1233 adult patients with NVUGIB who presented to the ED of a tertiary center between January 2022 and January 2025. Clinical and laboratory data were extracted from electronic records. Seven supervised ML algorithms—logistic regression, ridge regression, support vector machine, random forest, extreme gradient boosting (XGBoost), naïve Bayes, and artificial neural networks—were trained using six feature selection techniques generating 42 distinct models. Performance was assessed using AUROC, F1-score, sensitivity, specificity, and calibration metrics. Traditional scores (GBS, AIMS65, Rockall) were evaluated in parallel. Results: Among the cohort, 96 patients (7.8%) died within 30 days. The best-performing ML model (XGBoost with univariate feature selection) achieved an AUROC > 0.80 and F1-score of 0.909, significantly outperforming all traditional scores (highest AUROC: Rockall, 0.743; p < 0.001). ML models demonstrated higher sensitivity and specificity, with improved calibration. Key predictors consistently included age, comorbidities, hemodynamic parameters, and laboratory markers. The best-performing ML models demonstrated very high apparent AUROC values (up to 0.999 in internal analysis), substantially exceeding conventional scores. These results should be interpreted as apparent performance estimates, likely optimistic in the absence of external validation. Conclusions: While machine-learning models showed markedly higher apparent discrimination than conventional scores, these findings are based on a single-center retrospective dataset and require external multicenter validation before clinical implementation. Full article

Background and Objectives: The anesthetic nerve block test is a surgical technique that can assist in the differential diagnosis of forefoot pain. The MTP joint, enclosed by its capsule, may act as a sealed cavity with predictable contrast dispersion, whereas the IM space, lacking clear boundaries and containing bursae and the plantar digital nerve, favors diffuse spread. Due to the high rate of false positives in suspected cases of Morton’s neuroma with the anesthetic block current procedure in the intermetatarsal space, the aim of this study was to propose an alternative to the current procedure. Material and Methods: Six fresh cadaveric feet were used. Under ultrasound guidance, the 2nd–4th MTP joints received stepwise intra-articular injections of radiopaque contrast. The third common digital nerve was injected within the third intermetatarsal space. Standard radiographs were obtained to assess distribution and proximal spread. Results: A volume of 0.3 mL was sufficient to fully reach the intra-articular cavity and potentially induce effective localized anesthesia. When the third common digital plantar nerve was injected in an anatomically healthy region, the contrast medium showed a proximal diffusion pattern extending up to the mid-diaphyseal level of the third and fourth metatarsal bones. On radiographs, the intra-articular infiltration lines appear sharply demarcated, supporting the interpretation of the metatarsophalangeal joint as a sealed compartment. Conclusions: Low intra-articular anesthetic volumes may yield targeted effects, while Morton’s neuroma injections spread proximally, risking loss of diagnostic specificity; this technique may improve decision-making accuracy and reduce failures. Full article

Background/Objectives: Mycophenolic acid (MPA) is used as part of first-line combination immunosuppressive therapy for renal transplant recipients. Personalized dosing approaches are needed to balance efficacy and minimize toxicity due to the pharmacokinetic variability of the drug. In this study, population pharmacokinetic (PopPK) modeling and machine learning (ML) techniques are coupled to provide valuable insights into optimizing MPA therapy. Methods: Using data from 76 renal transplant patients, two PopPK models were developed to describe and predict MPA levels for two different formulations (enteric-coated mycophenolate sodium and mycophenolate mofetil). Covariate effects on drug clearance were assessed, and Monte Carlo simulations were used to evaluate exposure under normal and reduced clearance conditions. ML techniques, including principal component analysis (PCA) and ensemble tree models (bagging and boosting), were applied to identify predictive factors and explore associations between MPA plasma/saliva concentrations and the examined covariates. Results: Total daily dose and post-transplant time (PTP) were identified as key covariates affecting clearance. PCA highlighted MPA dose as the primary determinant of plasma levels, with urea and PTP also playing significant roles. Boosted tree analysis confirmed these findings, demonstrating strong predictive accuracy (R² > 0.91). Incorporating saliva MPA levels improved predictive performance, suggesting that saliva may be a complementary monitoring tool, although plasma monitoring remained superior. Simulations allowed exploring potential dosing adjustments for patients with reduced clearance. Conclusions: This study demonstrates the potential of integrating machine learning with population pharmacokinetic modeling to improve the understanding of MPA variability and support individualized dosing strategies in renal transplant recipients. The developed PopPK/ML models provide a methodological foundation for future research toward more personalized immunosuppressive therapy. Full article

This study presents a gradient boosting-based machine learning (ML) approach developed to predict cutting speed and feed rate in band sawing operations. The model was built using a dataset of 1701 experimental samples from three industrially common material types: AISI 304, CK45, and AISI 4140. Each sample was defined by key process parameters, namely, material type, a hardness range of 15–44 HRC, and a diameter range of 100–500 mm, with cutting speed and feed rate as target variables. Five ML models were examined and compared in this study, including linear regression (LR), support vector regression (SVR), random forest regression (RFR), least squares boosting (LSBoost), and extreme gradient boosting (XGBoost). Model training and validation were carried out using five-fold cross-validation. The results show that the XGBoost model offers the highest accuracy. For cutting speed estimation, the performance values of XGBoost are an RMSE of 0.213, an MAE of 0.140, an R² of 0.999, and an MAPE of 0.407%; and for feed rate estimation, an RMSE of 0.259, an MAE of 0.169, an R² of 0.999, and a MAPE of 1.14%. These results indicate that gradient-based ensemble methods capture the nonlinear behavior of cutting parameters more effectively than linear or kernel-driven techniques, providing a practical and robust approach for data-driven optimization in intelligent manufacturing. Full article

Carbon Fiber-Reinforced Polymer (CFRP) composites, widely used across industries, exhibit various damage mechanisms depending on the loading conditions applied. This study employs a structural health monitoring (SHM) approach to investigate the three primary failure modes, fiber breakage, matrix cracking, and delamination, in thermoset quasi-isotropic CFRPs subjected to quasi-static tensile loading until failure. Acoustic emission (AE) signals acquired from an experiment were leveraged to analyze and classify these real-time signals into the failure modes using machine learning (ML) techniques. Due to the extensive number of AE signals recorded during testing, manually classifying these failure mechanisms through waveform inspection was impractical. ML, alongside ensemble learning, algorithms were implemented to streamline the classification, making it more efficient, accurate, and reliable. Conventional AE parameters from the data acquisition system and feature extraction techniques applied to the recorded waveforms were implemented exclusively as classification features to investigate their reliability and accuracy in classifying failure modes in CFRPs. The classification models exhibited up to 99% accuracy, as depicted by evaluation metrics. Further studies, using cross-correlation techniques, ascertained the presence of fiber break events occurring in the bundles as the thermoset CFRP composite approached failure. These findings highlight the significance of integrating machine learning into SHM for the early detection of real-time damage and effective monitoring of residual life in composite materials. Full article

Background and Objectives: Diffuse idiopathic skeletal hyperostosis (DISH)-related spinal fractures with marked anterior distraction are highly unstable and pose substantial surgical challenges. The transdiscal screw for diffuse idiopathic skeletal hyperostosis (TSD) technique has been proposed to enhance fixation strength by penetrating adjacent vertebral endplates; however, its clinical utility in large-displacement cases remained unclear. Materials and Methods: In this retrospective study, we reviewed 21 patients with thoracolumbar DISH-related fractures and an anterior fracture gap ≥ 15 mm, who underwent posterior fixation between 2010 and 2024. 11 patients underwent TSD fixation (TSD group), and 10 underwent conventional fixation without bilateral TSD (control group). Results: The mean number of fused segments did not differ significantly between the groups (5.0 ± 1.4 vs. 5.0 ± 1.3, p = 0.43). Operative time was comparable (164 ± 57 vs. 168 ± 60 min, p = 0.90). Blood loss tended to be lower in the TSD group (306 ± 334 vs. 528 ± 658 mL, p = 0.33). For fracture-gap reduction, the TSD group improved from 17.4 ± 2.3 mm preoperatively to 13.8 ± 4.4 mm postoperatively and 2.0 ± 3.6 mm at final follow-up, while the control group showed less reduction (16.8 ± 2.2, 15.4 ± 1.4, and 7.0 ± 9.1 mm, respectively). Screw loosening occurred in three TSD patients and six controls (p = 0.13). All patients in the TSD group achieved bone union without reoperation, whereas four controls experienced implant backout, three required reoperation, and two failed to achieve bone union (p = 0.035). Conclusions: Posterior fixation using TSD provided reliable stability, maintained reduction, and reduced the risk of implant failure compared with conventional fixation in highly unstable DISH-related fractures with anterior distraction. Although larger prospective studies are needed, TSD may represent a valuable surgical option for this challenging patient population. Full article

Liquefied natural gas plays a crucial role in global energy transitions due to its high efficiency and low emissions, especially in long-distance transportation. However, the thermal management of LNG storage tanks remains a significant challenge due to temperature fluctuations, which impact both efficiency and safety. Traditional methods rely on thermodynamic models or computational fluid dynamics simulations but are computationally expensive and time-consuming. This study proposes a hybrid approach that integrates machine learning techniques with CFD data to predict temperature variations inside LNG storage tanks. Several ML models, including Random Forest, XGBoost, and deep learning-based models like CNN and TCN, were tested. Results indicate that CNN and TCN models offer the best performance in predicting temperature changes, showing superior accuracy and computational efficiency. This approach significantly enhances the real-time prediction capability, offering a promising solution for improving LNG tank thermal management, ensuring both operational safety and efficiency. Full article

Predicting the quality of freshwater, both surface and groundwater, is essential for the sustainable management of water resources. This study collected 1822 articles from the Scopus database (2000–2024) and filtered them using Topic Modeling to create the study corpus. The B-SLR analysis identified exponential growth in scientific publications since 2020, indicating that this field has reached a stage of maturity. The results showed that the predominant techniques for predicting water quality, both for surface and groundwater, fall into three main categories: (i) ensemble models, with Bagging and Boosting representing 43.07% and 25.91%, respectively, particularly random forest (RF), light gradient boosting machine (LightGBM), and extreme gradient boosting (XGB), along with their optimized variants; (ii) deep neural networks such as long short-term memory (LSTM) and convolutional neural network (CNN), which excel at modeling complex temporal dynamics; and (iii) traditional algorithms like artificial neural network (ANN), support vector machines (SVMs), and decision tree (DT), which remain widely used. Current trends point towards the use of hybrid and explainable architectures, with increased application of interpretability techniques. Emerging approaches such as Generative Adversarial Network (GAN) and Group Method of Data Handling (GMDH) for data-scarce contexts, Transfer Learning for knowledge reuse, and Transformer architectures that outperform LSTM in time series prediction tasks were also identified. Furthermore, the most studied water bodies (e.g., rivers, aquifers) and the most commonly used water quality indicators (e.g., WQI, EWQI, dissolved oxygen, nitrates) were identified. The B-SLR and Topic Modeling methodology provided a more robust, reproducible, and comprehensive overview of AI/ML/DL models for freshwater quality prediction, facilitating the identification of thematic patterns and research opportunities. Full article

Early detection and diagnosis of plant diseases is critical for ensuring global food security and sustainable agricultural practices. This review comprehensively examines latest advancements in crop disease risk prediction, onset detection through imaging techniques, machine learning (ML), deep learning (DL), and edge computing technologies. Traditional disease detection methods, which rely on visual inspections, are time-consuming, and often inaccurate. While chemical analyses are accurate, they can be time consuming and leave less flexibility to promptly implement remedial actions. In contrast, modern techniques such as hyperspectral and multispectral imaging, thermal imaging, and fluorescence imaging, among others can provide non-invasive and highly accurate solutions for identifying plant diseases at early stages. The integration of ML and DL models, including convolutional neural networks (CNNs) and transfer learning, has significantly improved disease classification and severity assessment. Furthermore, edge computing and the Internet of Things (IoT) facilitate real-time disease monitoring by processing and communicating data directly in/from the field, reducing latency and reliance on in-house as well as centralized cloud computing. Despite these advancements, challenges remain in terms of multimodal dataset standardization, integration of individual technologies of sensing, data processing, communication, and decision-making to provide a complete end-to-end solution for practical implementations. In addition, robustness of such technologies in varying field conditions, and affordability has also not been reviewed. To this end, this review paper focuses on broad areas of sensing, computing, and communication systems to outline the transformative potential of end-to-end solutions for effective implementations towards crop disease management in modern agricultural systems. Foundation of this review also highlights critical potential for integrating AI-driven disease detection and predictive models capable of analyzing multimodal data of environmental factors such as temperature and humidity, as well as visible-range and thermal imagery information for early disease diagnosis and timely management. Future research should focus on developing autonomous end-to-end disease monitoring systems that incorporate these technologies, fostering comprehensive precision agriculture and sustainable crop production. Full article

Background: Massive intraoperative blood loss (IBL) is a serious complication in complex spine surgeries such as deformity correction, multilevel fusion, tumor resection, and revision procedures. While no strict definition exists, blood loss exceeding 1500 mL or 20% of estimated blood volume is generally considered clinically significant. Excessive bleeding increases the risk of hemodynamic instability, transfusion-related complications, postoperative infection, and prolonged hospitalization. Methods: This narrative review summarizes the current understanding of the incidence, risk factors, anatomical vulnerabilities, and evidence-based strategies for managing IBL in spine surgery through comprehensive literature analysis of recent studies and clinical guidelines. Results: Key risk factors include patient characteristics (anemia, obesity, advanced age, medication use), surgical variables (multilevel instrumentation, revision status, operative time), and pathological conditions (hypervascular tumors, severe deformity). Perioperative medication management is critical, requiring discontinuation of NSAIDs (5–7 days), antiplatelet agents (5–7 days), and NOACs (48–72 h) preoperatively to minimize bleeding risk. The thoracolumbar junction and hypervascular spinal lesions are especially prone to bleeding due to dense vascular anatomy. Evidence-based management strategies include comprehensive preoperative optimization, intraoperative hemostatic techniques, antifibrinolytic agents, topical hemostatic products, cell salvage technology, and structured transfusion protocols. Conclusions: Effective management of massive IBL requires a multimodal approach combining preoperative risk assessment and medication optimization, intraoperative hemostatic strategies including tranexamic acid administration, advanced monitoring techniques, and coordinated transfusion protocols. Particular attention to perioperative management of anticoagulant and antiplatelet medications is essential for bleeding risk mitigation. Understanding patient-specific risk factors, surgical complexity, and anatomical considerations enables surgeons to implement targeted prevention and management strategies, ultimately improving patient outcomes and reducing complications in high-risk spine surgery procedures. Full article

Bone drilling is a widely encountered process in orthopedic surgeries and keyhole neuro surgeries. We are developing a sensor-integrated smart end-effector for drilling for robotic surgical applications. In manual surgeries, surgeons assess tool wear based on experience and force perception. In this work, we propose a machine-learning (ML)-based tool condition monitoring system based on multi-sensor data to preempt excessive tool wear during drilling in robotic surgery. Real-time data is acquired from the six-component force sensor of a collaborative arm along with the data from the temperature and multi-axis vibration sensor mounted on the bone specimen being drilled upon. Raw data from the sensors may have noises and outliers. Signal processing in the time- and frequency-domain are used for denoising as well as to obtain additional features to be derived from the raw sensory data. This paper addresses the challenging problem of identification of the most suitable ML algorithm and the most suitable features to be used as inputs to the algorithm. While dozens of features and innumerable machine learning and deep learning models are available, this paper addresses the problem of selecting the most relevant features, the most relevant AI models, and the optimal hyperparameters to be used in the AI model to provide accurate prediction on the tool condition. A unique framework is proposed for classifying tool wear that combines machine learning-based modeling with multi-sensor data. From the raw sensory data that contains only a handful of features, a number of additional features are derived using frequency-domain techniques and statistical measures. Using feature engineering, we arrived at a total of 60 features from time-domain, frequency-domain, and interaction-based metrics. Such additional features help in improving its predictive capabilities but make the training and prediction complicated and time-consuming. Using a sequence of techniques such as variance thresholding, correlation filtering, ANOVA F-test, and SHAP analysis, the number of features was reduced from 60 to the 4 features that will be most effective in real-time tool condition prediction. In contrast to previous studies that only examine a small number of machine learning models, our approach systematically evaluates a wide range of machine learning and deep learning architectures. The performances of 47 classical ML models and 6 deep learning (DL) architectures were analyzed using the set of the four features identified as most suitable. The Extra Trees Classifier (an ML model) and the one-dimensional Convolutional Neural Network (1D CNN) exhibited the best prediction accuracy among the models studied. Using real-time data, these models monitored the drilling tool condition in real-time to classify the tool wear into three categories of slight, moderate, and severe. Full article

Slope failures represent one of the most serious geotechnical hazards, which can have severe consequences for personnel, equipment, infrastructure, and other aspects of a mining operation. Deterministic and stochastic conventional methods of slope stability analysis are useful; however, some limitations in applicability may arise due to the inherent anisotropy of rock mass properties and rock mass interactions. In recent years, Machine Learning (ML) techniques have become powerful tools for improving prediction and risk assessment in slope stability analysis. This review provides a comprehensive overview of ML applications for analyzing slope stability and delves into the performance of each technique as well as the interrelationship between the geotechnical parameters of the rock mass. Supervised learning methods such as decision trees, support vector machines, random forests, gradient boosting, and neural networks have been applied by different authors to predict the safety factor and classify slopes. Unsupervised learning techniques such as clustering and Gaussian mixture models have also been applied to identify hidden patterns. The objective of this manuscript is to consolidate existing work by highlighting the advantages and limitations of different ML techniques, while identifying gaps that should be analyzed in future research. Full article

Background/Objectives: Chronic kidney disease (CKD) is a progressive condition that affects the body’s ability to remove waste and regulate fluid and electrolytes. Early detection is crucial for delaying disease progression and initiating timely interventions. Machine learning (ML) techniques have emerged as powerful tools for automating disease diagnosis and prognosis. This study aims to evaluate the predictive performance of individual and ensemble ML algorithms for the early classification of CKD. Methods: A clinically annotated dataset was utilized to categorize patients into CKD and non-CKD groups. The models investigated included Logistic Regression, Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), Ridge Classifier, Naïve Bayes, K-Nearest Neighbors (KNN), Decision Tree (DT), Random Forest (RF), Support Vector Machine (SVM), and Ensemble learning strategies. A systematic preprocessing pipeline was implemented, and model performance was assessed using accuracy, precision, recall, F1 score, and area under the receiver operating characteristic curve (AUC). Results: The empirical findings reveal that ML-based classifiers achieved high predictive accuracy in CKD detection. Ensemble learning methods outperformed individual models in terms of robustness and generalization, indicating their potential in clinical decision-making contexts. Conclusions: The study demonstrates the efficacy of ML-based frameworks for early CKD prediction, offering a scalable, interpretable, and accurate clinical decision support approach. The proposed methodology supports timely diagnosis and can assist healthcare professionals in improving patient outcomes. Full article

Background and Objectives: While transurethral resection of the prostate remains the gold standard for surgical treatment of benign prostatic hyperplasia, anatomical endoscopic enucleation of the prostate provides a safe, durable, and size-independent alternative. Our study introduces a specific technical innovation, i.e., en bloc bipolar prostate enucleation performed exclusively via sheath-tip mechanical dissection without the use of a dedicated enucleation loop, combined with the mushroom technique and early apical release. Materials and Methods: Between January 2018 and May 2023, 252 patients with prostate volumes > 30 mL and significant lower urinary tract symptoms underwent en bloc bipolar prostate enucleation via the mushroom technique with early apical release. Data were retrospectively evaluated to assess perioperative results, postoperative outcomes, and complications. Results: The median age of the cohort was 70 (65–76) years, with a median prostate volume of 60 (40–88.5) mL. The median operative time was 40 (30–70) min, and the median weight of enucleated tissue was 34 (16.5–60) g. Significant improvements were observed in the International Prostate Symptom score, Quality of Life score, maximum flow rate, average flow rate, and postvoid residual urine at 12 months (p < 0.001). The rate of transient stress urinary incontinence decreased from 19.44% at 1 month to 2.38% at 12 months. Conclusions: En bloc bipolar prostate enucleation using the mushroom technique is a safe and effective treatment for benign prostatic hyperplasia, yielding significant improvements in urinary symptoms and flow rates, with a manageable complication profile. Further multicenter studies are needed to confirm these findings. Full article

The integration of machine learning (ML) techniques with unmanned aerial vehicle (UAV) imagery holds strong potential for improving yield prediction in agriculture. However, few studies have combined biophysical field variables with UAV-derived spectral data, particularly under conditions of limited sample size. This study evaluated the performance of different ML algorithms in predicting Arabica coffee (Coffea arabica) yield using field-based biophysical measurements and spectral variables extracted from multispectral UAV imagery. The research was conducted over two crop seasons (2020/2021 and 2021/2022) in a 1.2-hectare experimental plot in southeastern Brazil. Three modeling scenarios were tested with Random Forest, Gradient Boosting, K-Nearest Neighbors, Multilayer Perceptron, and Decision Tree algorithms, using Leave-One-Out cross-validation. Results varied considerably across seasons and scenarios. KNN performed best with raw data, while Gradient Boosting was more stable after variable selection and synthetic data augmentation with SMOTE. Nevertheless, limitations such as small sample size, seasonal variability, and overfitting, particularly with synthetic data, affected overall performance. Despite these challenges, this study demonstrates that integrating UAV-derived spectral data with ML can support yield estimation, especially when variable selection and phenological context are carefully addressed. Full article