Search Results (4,835)

(1) Background: Invasive micropapillary carcinoma (IMPC) is a rare, aggressive breast cancer subtype marked by high lymph node metastasis rates. While Oncotype DX recurrence score (RS) offers prognostic information for patients with hormone-receptor-positive (HR+) breast cancer, its utility in IMPC—a histology with distinct biologic behavior—remains unvalidated. This study evaluates whether Oncotype DX offers prognostic information with respect to overall survival (OS) in non-metastatic, early-stage patients with IMPC of the breast. (2) Methods: The National Cancer Database (2004–2020) was queried to select for women with ER+/HER2−, T1-T2N0-N1 IMPC who underwent Oncotype DX testing and received no neoadjuvant therapy. Patients were stratified by RS: low (≤11), intermediate (12–25), and high (>25). Kaplan–Meier survival curves and log-rank tests compared 5-year OS between groups. Multivariable Cox proportional hazards models assessed RS as an independent predictor, adjusting for age, race, comorbidities, grade, radiation, and insurance status. (3) Results: A total of 1325 women met the selection criteria. The cohort demonstrated significant survival disparities by RS (log-rank p = 0.017). Five-year OS rates were 97.5%, 97.5%, and 93.7% for low, intermediate, and high-risk patients, respectively. Adjusted multivariate analysis confirmed RS as an independent prognosticator: low (HR = 0.31, 95% CI: 0.15–0.75) and intermediate (HR = 0.32, 95% CI: 0.15–0.75) scores correlated with reduced mortality versus high RS. Omission of radiation therapy (HR = 2.68, 95% CI: 1.05–6.86) and higher comorbidity burden (0 comorbidities vs. ≥2: HR = 0.25, 95% CI: 0.10–0.61) were significantly associated with worse survival. (4) Conclusions: Oncotype DX is predictive for OS in IMPC, with high RS (>25) portending poorer outcomes. The survival detriment associated with RT omission aligns with prior studies demonstrating RT benefit in higher-risk cohorts. These findings validate RS as a prognostic tool in IMPC and underscore its potential to refine adjuvant therapy, particularly RT utilization. Future studies should explore RS-driven treatment personalization in IMPC, including comorbidity management and adjuvant radiation to improve outcomes in this distinct patient population. Full article

Background: Atherogenic dyslipidemia is defined by the coexistence of high triglyceride concentrations, low levels of high-density lipoprotein cholesterol (HDL-C), and an excess of small, dense particles of low-density lipoprotein cholesterol (LDL-C). This lipid profile is strongly associated with an increased burden of cardiovascular disease and represents a leading cause of global morbidity and mortality. To better capture this risk, composite lipid ratios—including total cholesterol to HDL-C (TC/HDL-C), LDL-C to HDL-C (LDL-C/HDL-C), triglycerides to HDL-C (TG/HDL-C), and the atherogenic dyslipidemia index (AD)—have emerged as robust markers of cardiometabolic health, frequently demonstrating superior predictive capacity compared with isolated lipid measures. Despite extensive evidence linking these ratios to cardiovascular disease, few large-scale studies have examined their association with sociodemographic characteristics, lifestyle behaviors, and social isolation in working populations. Methods: We conducted a cross-sectional analysis of a large occupational cohort of Spanish workers evaluated between January 2021 and December 2024. Anthropometric, biochemical, and sociodemographic data were collected through standardized clinical protocols. Indices of atherogenic risk—namely the ratios TC/HDL-C, LDL-C/HDL-C, TG/HDL-C, and the atherogenic dyslipidemia index (AD)—were derived from fasting lipid measurements. The assessment of lifestyle factors included tobacco use, physical activity evaluated through the International Physical Activity Questionnaire (IPAQ), adherence to the Mediterranean dietary pattern using the MEDAS questionnaire, and perceived social isolation measured by the Lubben Social Network Scale. Socioeconomic classification was established following the criteria proposed by the Spanish Society of Epidemiology. Logistic regression models were fitted to identify factors independently associated with moderate-to-high risk for each lipid indicator, adjusting for potential confounders. Results: A total of 117,298 workers (71,384 men and 45,914 women) were included. Men showed significantly higher odds of elevated TG/HDL-C (OR 4.22, 95% CI 3.70–4.75) and AD (OR 2.95, 95% CI 2.70–3.21) compared with women, whereas LDL-C/HDL-C ratios were lower (OR 0.86, 95% CI 0.83–0.89). Advancing age was positively associated with all lipid ratios, with the highest risk observed in participants aged 60–69 years. Lower social class, smoking, physical inactivity, poor adherence to the Mediterranean diet, and low social isolation scores were consistently linked to higher atherogenic risk. Physical inactivity showed the strongest associations across all indicators, with ORs ranging from 3.54 for TC/HDL-C to 7.12 for AD. Conclusions: Atherogenic dyslipidemia and elevated lipid ratios are strongly associated with male sex, older age, lower socioeconomic status, unhealthy lifestyle behaviors, and reduced social integration among Spanish workers. These findings highlight the importance of workplace-based cardiovascular risk screening and targeted prevention strategies, particularly in high-risk subgroups. Interventions to promote physical activity, healthy dietary patterns, and social connectedness may contribute to lowering atherogenic risk in occupational settings. Full article

Accurate prediction of earthquake—induced building damage is essential for timely disaster response and effective risk mitigation. This study explores a machine learning (ML)-based classification approach using data from the 2015 Gorkha, Nepal earthquake, with a specific focus on reinforced concrete (RC) structures. The original dataset from the 2015 Nepal earthquake contained 762,094 building entries across 127 variables describing structural, functional, and contextual characteristics. Three ensemble ML modelsGradient Boosting Machine (GBM), Extreme Gradient Boosting (XGBoost), and Light Gradient Boosting Machine (LightGBM) were trained and tested on both the full dataset and a filtered RC-only subset. Two target variables were considered: a three-class variable (damage_class) and the original five-level damage grade (damage_grade). To address class imbalance, oversampling and undersampling techniques were applied, and model performance was evaluated using accuracy and F1 scores. The results showed that LightGBM consistently outperformed the other models, especially when oversampling was applied. For the RC dataset, LightGBM achieved up to 98% accuracy for damage_class and 93% accuracy for damage_grade, along with high F1 scores ranging between 0.84 and 1.00 across all classes. Feature importance analysis revealed that structural characteristics such as building area, age, and height were the most influential predictors of damage. These findings highlight the value of building-type-specific modeling combined with class balancing techniques to improve the reliability and generalizability of ML-based earthquake damage prediction. Full article

Background and Objectives: Sarcopenia is a progressive loss of skeletal muscle mass and function in elderly adults, posing a significant risk of frailty, falls, and morbidity. The current study designs and evaluates SarcoNet, a novel artificial neural network (ANN)-based classification framework developed in order to classify Sarcopenic from non-Sarcopenic subjects using a comprehensive real-time dataset. Methods: This pilot study involved 30 subjects, who were divided into Sarcopenic and non-Sarcopenic groups based on physician assessment. The collected dataset consists of thirty-one clinical parameters like skeletal muscle mass, which is collected using various equipment such as Body Composition Analyser, along with ten kinetic features which are derived from video-based gait analysis of joint angles obtained during walking on three terrain types such as slope, steps, and parallel path. The performance of the designed ANN-based SarcoNet was benchmarked against the traditional machine learning classifiers utilised including Support Vector Machine (SVM), k-Nearest Neighbours (k-NN), and Random Forest (RF), as well as hard and soft voting ensemble classifiers. Results: SarcoNet achieved the highest overall classification accuracy of about 94%, with a specificity and precision of about 100%, an F1-score of about 92.4%, and an AUC of 0.94, outperforming all other models. The incorporation of lower-limb joint kinetics such as knee flexion, extension, ankle plantarflexion and dorsiflexion significantly enhanced predictive capability of the model and thus reflecting the functional deterioration characteristic of muscles in Sarcopenia. Conclusions: SarcoNet provides a promising AI-driven solution in Sarcopenia diagnosis, especially in low-resource healthcare settings. Future work will focus on improving the dataset, validating the model across diverse populations, and incorporating explainable AI to improve clinical adoption. Full article

This paper introduces the Gesture-Based Physical Stability Classification and Rehabilitation System (GPSCRS), a low-cost, non-invasive solution for evaluating physical stability using an Arduino microcontroller and the DFRobot Gesture and Touch sensor. The system quantifies movement smoothness, consistency, and speed by analyzing “up” and “down” hand gestures over a fixed period, generating a Physical Stability Index (PSI) as a single metric to represent an individual’s stability. The system focuses on a temporal analysis of gesture patterns while incorporating placeholders for speed scores to demonstrate its potential for a comprehensive stability assessment. The performance of various machine learning and deep learning models for gesture-based classification is evaluated, with neural network architectures such as Transformer, CNN, and KAN achieving perfect scores in recall, accuracy, precision, and F1-score. Traditional machine learning models such as XGBoost show strong results, offering a balance between computational efficiency and accuracy. The choice of model depends on specific application requirements, including real-time constraints and available resources. The preliminary experimental results indicate that the proposed GPSCRS can effectively detect changes in stability under real-time conditions, highlighting its potential for use in remote health monitoring, fall prevention, and rehabilitation scenarios. By providing a quantitative measure of stability, the system enables early risk identification and supports tailored interventions for improved mobility and quality of life. Full article

Background: Intensive Care Unit (ICU) readmission and in-hospital mortality are critical indicators of patient outcomes following ICU discharge. Patients readmitted to the ICU often face worse prognosis, higher healthcare costs, and prolonged hospital stays. Identifying high-risk patients is essential for optimizing post-ICU care and resource allocation. Methods: This two-phase study included the following: (1) a retrospective analysis of ICU survivors in a mixed medical–surgical ICU to identify risk factors associated with ICU readmission and in-hospital mortality, and (2) a prospective validation of a newly developed predictive model: the Worse Outcome Score (WOScore). Data collected included demographics, ICU admission characteristics, severity scores (SAPS II, SAPS III, APACHE II, SOFA), interventions, complications and discharge parameters. Results: Among 1.190 ICU survivors, 126 (10.6%) were readmitted to the ICU, and 192 (16.1%) died in hospital after ICU discharge. Key risk factors for ICU readmission included Diabetes Mellitus, SAPS III on admission, and ICU-acquired infections (Ventilator-Associated Pneumonia (VAP) and Catheter-Related Bloodstream Infection, (CRBSI)). Predictors of in-hospital mortality were identified: medical admission, high SAPS III score, high lactate level on ICU admission, tracheostomy, reduced GCS at discharge, blood transfusion, CRBSI, and Acute Kidney Injury (AKI) during ICU stay. The WOScore, developed based on the results above, demonstrated strong predictive ability (AUC: 0.845 derivation, 0.886 validation). A cut-off of 20 distinguished high-risk patients (sensitivity: 88.1%, specificity: 73.0%). Conclusions: ICU readmission and in-hospital mortality are influenced by patient severity, underlying comorbidities, and ICU-related complications. The WOScore provides an effective, easy-to-use risk stratification tool that can guide clinicians in identifying high-risk patients at ICU discharge and guide post-ICU interventions, potentially improving patients’ outcomes and optimizing resource allocation. Further multi-center studies are necessary to validate the model in diverse healthcare settings. Full article

Background/Objectives: The attainment of a stone-free (SF) condition is a fundamental indicator of successful outcomes after flexible ureteroscopy (fURS) for urinary stone disease. External confirmations of preoperative scores remain limited. We externally validated the T.O.HO. and S.T.O.N.E. scores in an independent Japanese cohort and examined calibration, decision curve utility, and threshold-guided use to support personalized planning. Methods: We retrospectively analyzed 361 consecutive patients treated with fURS from March 2018 to August 2023. Postoperative SF status was defined as the absence of residual calculi greater than 2 mm on non-contrast computed tomography performed within three months of surgery. Independent determinants of SF were identified using multivariable logistic regression, predictive performance was quantified by receiver operating characteristic analyses with DeLong’s test, and model calibration and decision curve analysis were additionally assessed. Results: Among the 361 patients, 255 (70.6%) achieved an SF state. A larger stone diameter, the presence of lower-pole calculi, and preoperative pyuria (positive urine WBC) were significant independent predictors of residual fragments. T.O.HO. demonstrated superior discrimination (AUC 0.86) compared with S.T.O.N.E. (AUC 0.77; p < 0.01) and surpassed individual predictors. Both scores showed acceptable calibration. Decision curve analysis demonstrated higher net benefit for T.O.HO. across clinically relevant thresholds. We provide clinically useful cut-offs (e.g., T.O.HO. ≤5: high SF probability; 6: trade-off discussion; ≥7: higher residual risk) to align actions with patient priorities. Conclusions: Beyond discrimination, a calibrated, threshold-aware use of T.O.HO. enables personalized preoperative counseling and shared decision-making, potentially reducing unnecessary staging and enhancing routine fURS planning. Full article

Background: Type 2 diabetes mellitus (T2DM) is a global health concern driven by aging, lifestyle, and socio-economic disparities. Early detection is key, with tools like FINDRISC, QDScore, and CANRISK providing non-invasive screening. Yet, the combined effects of sociodemographic factors, healthy habits, and perceived quality of life on diabetes risk remain insufficiently studied in working populations. Objectives: To evaluate the association between sociodemographic variables, lifestyle habits (smoking, physical activity, adherence to the Mediterranean diet), and health-related quality of life (HRQoL) with the risk of developing type 2 diabetes, using three validated screening tools in a large cohort of Spanish workers. Methods: A cross-sectional study was conducted among 100,014 Spanish workers aged 18 to 69 years who underwent standardized medical evaluations between January 2021 and December 2023. Diabetes risk was assessed using the FINDRISC, QDScore, and CANRISK tools. Lifestyle variables and HRQoL (measured via the SF-12 questionnaire) were evaluated through validated instruments. Multivariate logistic regression models were used to examine the association of independent variables with moderate-to-high diabetes risk according to each score. Results: Among the strongest predictors, poor adherence to a Mediterranean diet (OR = 1.45, 95% CI: 1.32–1.58) and low physical activity (OR = 1.39, 95% CI: 1.27–1.52) were independently associated with higher diabetes risk. Poor HRQoL was also significant (OR = 1.33, 95% CI: 1.22–1.47). Conclusions: Sociodemographic factors, lifestyle behaviors, and perceived health status are independently associated with increased type 2 diabetes risk in Spanish workers. The integration of HRQoL assessments into occupational health surveillance may enhance early identification of at-risk individuals and guide tailored prevention strategies. Full article

Background and Objectives: To date, several machine learning (ML) prognostic prediction models have been investigated for patients with acute myocardial infarction (AMI). However, few studies have compared the prognostic performance of ML techniques in AMI patients who underwent percutaneous coronary intervention (PCI). We sought to compare the prognostic performance among various machine learning techniques to determine which one showed the best prediction ability. Materials and Methods: Using data from the large, multicenter COREA-AMI registry, this study analyzed 10,172 patients to predict major adverse cardiac events (MACEs) at 1 and 5 years. MACE was defined as a composite of cardiac death, myocardial infarction, or cerebrovascular accident. Results: Compared with the four other ML techniques and traditional logistic regression, the random forest (RF) model consistently demonstrated the highest predictive performance. At 5 years, the RF model achieved a superior area under the curve (AUC) of 0.822, an accuracy of 0.804, and an F1 score of 0.870. To ensure clinical interpretability, a SHapley Additive exPlanations analysis was performed on the RF model. It identified key independent predictors for MACEs. The top nonmodifiable predictors included age, renal function, and left ventricular ejection fraction, whereas modifiable risk factors included dual antiplatelet therapy, statin therapy, angiotensin-converting enzyme inhibitor/angiotensin receptor blocker therapy, and adherence to these optimal medical therapy. Conclusions: In this real-world patient cohort, the RF model provided modest improvements in long-term risk stratification, and our findings highlight the continuing importance of guideline-directed medical therapy in determining patient prognosis. Full article

In pharmaceutical manufacturing, the reliability of machines and utility assets is critical to ensuring product quality, regulatory compliance, and uninterrupted operations. Traditional Risk-Based Maintenance (RBM) models quantify asset criticality using the Risk Priority Number (RPN), calculated from the probability and impact of failure alongside detectability. However, these models often neglect the practical challenges involved in diagnosing and resolving equipment issues, particularly in GMP-regulated environments. This study proposes an enhanced RPN framework that replaces the conventional detectability component with Maintainability Complexity (MC), quantified through two practical indicators: Ease of Diagnosis (ED) and Ease of Resolution (ER). Thirteen Key Performance Indicators (KPIs) were developed to assess Probability, Impact, and MC across 185 pharmaceutical utility assets. To enable objective risk stratification, Jenks Natural Breaks Optimization was applied to group assets into Low, Medium, and High risk tiers. Both multiplicative and normalized averaging methods were tested for score aggregation, allowing comparative analysis of their impact on prioritization outcomes. The enhanced model produced stronger alignment with operational realities, enabling more accurate asset classification and maintenance scheduling. A 3D risk matrix was introduced to translate scores into proactive strategies, offering traceability and digital compatibility with Computerized Maintenance Management Systems (CMMS). This framework provides a practical, auditable, and scalable approach to maintenance planning, supporting Industry 4.0 readiness in pharmaceutical operations. Full article

We propose an edge-enhanced federated learning framework for real-time itinerary optimization in elderly oriented adventure tourism, addressing the critical need for adaptive scheduling that balances activity intensity with health constraints. The system integrates lightweight convolutional neural networks with a priority-based scheduling algorithm, processing participant profiles and real-time biometric data through a decentralized computation model to enable dynamic adjustments. A modified Hungarian algorithm incorporates physical exertion scores, temporal proximity weights, and health risk factors, then optimizes activity assignments while respecting physiological recovery requirements. The federated learning architecture operates across distributed edge nodes, preserving data privacy through localized model training and periodic global aggregation. Furthermore, the framework interfaces with transportation systems and medical monitoring infrastructure, automatically triggering itinerary modifications when vital sign anomalies exceed adaptive thresholds. Implemented on NVIDIA Jetson AGX Orin modules, the system achieves 300 ms end-to-end latency for real-time schedule updates, meeting stringent safety requirements for elderly participants. The proposed method demonstrates significant improvements over conventional itinerary planners through its edge computing efficiency and personalized adaptation capabilities, particularly in handling the latency-sensitive demands of intensive tourism scenarios. Experimental results show robust performance across diverse participant profiles and activity types, confirming the system’s practical viability for real-world deployment in elderly adventure tourism operations. Full article

In response to the problems of rough modeling and insufficient coverage of edge interaction scenarios in autonomous driving tests, this paper proposes a scene generation method based on interaction coding. The method constructs a hierarchical parameter system of function–logic–specific scene, uses the time difference of arrival at interaction points (TTC_diff) to determine the critical state of interaction, and realizes the efficient generation and iterative optimization of high-risk scenes. Taking the unprotected left turn at the signal intersection of urban roads as an example, the interaction coding combination is determined in combination with real traffic data, the test scene compatible with OpenSCENARIO is generated, and CARLA0.9.15 is called for test verification. The results show that the interaction intensity is significantly negatively correlated with the trustworthiness score (−0.815), the generated scene has high coverage, and both safety and challenge are taken into account. Compared with the simulated annealing method, the method in this paper performs better in terms of iteration efficiency, scene difficulty control, and score stability, which provides an efficient and reliable test strategy for the trustworthiness evaluation of autonomous driving. Full article

As artificial intelligence transitions from conversational agents to autonomous actors in high-stakes environments, a critical gap emerges: how to ensure AI prioritizes human safety when its core objectives conflict with human well-being. Current safety benchmarks focus on harmful content, not behavioral alignment during instrumental goal conflicts. To address this, we introduce PacifAIst, a benchmark of 700 scenarios testing self-preservation, resource acquisition, and deception. We evaluated eight state-of-the-art large language models, revealing a significant performance hierarchy. Google’s Gemini 2.5 Flash demonstrated the strongest human-centric alignment (90.31%), while the highly anticipated GPT-5 scored lowest (79.49%), indicating potential risks. These findings establish an urgent need to shift the focus of AI safety evaluation from what models say to what they would do, ensuring that autonomous systems are not just helpful in theory but are provably safe in practice. Full article

Immunogenic cell death (ICD) is a specialized form of cell death that triggers antitumor immune responses. In tumors, ICD promotes the release of tumor-associated and tumor-specific antigens, thereby reshaping the immune microenvironment, restoring antitumor immunity, and facilitating tumor eradication. However, the regulatory mechanisms of ICD and its immunological effects vary across tumor types, and a comprehensive understanding remains limited. We systematically analyzed the expression of 34 ICD-related regulatory genes across 33 tumor types. Differential expression at the RNA, copy number variation (CNV), and DNA methylation levels was assessed in relation to clinical features. Associations between patient survival and RNA expression, CNVs, single-nucleotide variations (SNVs), and methylation were evaluated. Patients were stratified into immunological subtypes and further divided into high- and low-risk groups based on optimal prognostic models built using a machine learning framework. We explored the relationships between ICD-related genes and immune cell infiltration, stemness, heterogeneity, immune scores, immune checkpoint and regulatory genes, and subtype-specific expression patterns. Moreover, we examined the influence of immunotherapy and anticancer immune responses, applied three machine learning algorithms to identify prognostic biomarkers, and performed drug prediction and molecular docking analyses to nominate therapeutic targets. ICD-related genes were predominantly overexpressed in ESCA, GBM, KIRC, LGG, PAAD, and STAD. RNA expression of most ICD-related genes was associated with poor prognosis, while DNA methylation of these genes showed significant survival correlations in LGG and UVM. Prognostic models were successfully established for 18 cancer types, revealing intrinsic immune regulatory mechanisms of ICD-related genes. Machine learning identified several key prognostic biomarkers across cancers, among which NT5E emerged as a predictive biomarker in head and neck squamous cell carcinoma (HNSC), mediating tumor–immune interactions through multiple ligand–receptor pairs. This study provides a comprehensive view of ICD-related genes across cancers, identifies NT5E as a potential biomarker in HNSC, and highlights novel targets for predicting immunotherapy response and improving clinical outcomes in cancer patients. Full article

Microgrids are becoming an innovative piece of modern energy systems as they provide locally sourced and resilient energy opportunities and enable efficient energy sourcing. However, microgrid operations can be greatly affected by sudden environmental changes, deviating demand, and unexpected outages. In particular, extreme climatic events expose the vulnerability of microgrid infrastructure and resilience, often leading to increased risk of system-wide outages. Thus, successful microgrid operation relies on timely and accurate outage predictions. This research proposes a data-driven machine learning framework for the optimized operation of a microgrid and predictive outage detection using a Recurrent Neural Network–Long Short-Term Memory (RNN-LSTM) architecture that reflects inherent temporal modeling methods. A time-aware embedding and masking strategy is employed to handle categorical and sparse temporal features, while mutual information-based feature selection ensures only the most relevant and interpretable inputs are retained for prediction. Moreover, the model addresses the challenges of experiencing rapid power fluctuations by looking at long-term learning dependency aspects within historical and real-time data observation streams. Two datasets are utilized: a locally developed real-time dataset collected from a 5 MW microgrid of Maple Cement Factory in Mianwali and a 15-year national power outage dataset obtained from Kaggle. Both datasets went through intensive preprocessing, normalization, and tokenization to transform raw readings into machine-readable sequences. The suggested approach attained an accuracy of 86.52% on the real-time dataset and 84.19% on the Kaggle dataset, outperforming conventional models in detecting sequential outage patterns. It also achieved a precision of 86%, a recall of 86.20%, and an F1-score of 86.12%, surpassing the performance of other models such as CNN, XGBoost, SVM, and various static classifiers. In contrast to these traditional approaches, the RNN-LSTM’s ability to leverage temporal context makes it a more effective and intelligent choice for real-time outage prediction and microgrid optimization. Full article