Search Results (672)

Background/Objectives: The human brain is responsible for controlling various physiological functions, and hence, the presence of tumors in the brain is a major concern in the medical field. The correct identification and categorization of tumors in the brain using Magnetic Resonance Imaging (MRI) is a major requirement for the diagnosis and treatment of a tumor. The proposed research will focus on designing a CNN model that is optimized for tumor image classification. Methods: This research proposes an optimized CNN model featuring strategically placed dropout layers and hyperparameter optimization. This study uses a dataset of 640 MRI scans (320 tumor and 320 non-tumor) collected from a private hospital in Saudi Arabia. The proposed method utilizes a learning rate of 0.001 in combination with the Adam optimizer to ensure stable and efficient convergence. Its performance was benchmarked against established architectures, including VGG-19, Inception V3, ResNet-10, and ResNet-50, with evaluation based on classification accuracy and computational cost. Results: The experimental results show that the optimized CNN proposed in this work performs much better than the deeper architectures. The network reached a maximum training accuracy of 97.77% and a final test accuracy of 95.35% with a small test loss of 0.2223. The test accuracy of the optimized VGG-19 and Inception V3 networks was much lower, with a training time per epoch that was several orders of magnitude higher. The validation stability of the proposed network was high (92.25% to 95.35%) during the final stages of training. Conclusions: The conclusion drawn from this study is that hyperparameter optimization and strategic regularization are more advantageous for tumor classification using MRI images than the mere depth of the model. The accuracy of 95.35% with low computational complexity makes this lightweight CNN model a feasible solution for real-time applications. Full article

Background/Objectives: Human papillomavirus (HPV) infection is a major concern in public health, given its role as a persistent sexually transmitted infection and a causative agent of non-cancerous and cancerous lesions (neoplasms). The increasing infection rates observed in recent years underscore the need for effective public health measures to address this issue. The objective of this study is to describe the challenges and the results of conducting vaccination campaigns within a university setting and its impact on the HPV vaccination rate. Methods: A multifaceted approach was adopted, entailing the implementation of two distinct interventions. Following the promotional and educational online campaign (described elsewhere), vaccination delivery took place from November 2024 to July 2025 in the university campus and in three university hospitals in Milan. Overall and covariate-specific drop-out rate is calculated; significance is tested through a chi-square test of homogeneity between the population that completed less than three doses vs. those who completed the full cycle. Overall and vaccine-specific vaccination proportion is reported. Results: The vaccination rate for first doses reached 92% of available appointments, with a slight female majority (50.9%) and the 23–26 age as the most represented group (47%). The most represented nationality was Italian (58.4%), followed by Iranian (26.5%). Regarding the vaccination sites, the university venue recorded the highest rates in terms of both vaccines booked (56.4%) and vaccines administered (64.7%). With a net loss in follow up, consistent with WHO data, the three-dose HPV vaccination campaign was completed by 82.5% of participants. A chi-squared test of homogeneity revealed significant differences in age distribution between vaccination groups, χ² (3) = 347.78, p < 0.001, Cramér’s V = 0.457. Participants who received only one dose were predominantly younger (17–22 years: 71.1% vs. 19.0%, difference = 52.1 percentage points, 95% CI [46.6, 57.7]). Meanwhile, a catch-up strategy raised interest on other crucial vaccinations. Conclusions: The findings pertaining to the vaccination rate underscore the heightened awareness among young adults concerning the HPV vaccine. They further substantiate the efficacy of the integrated strategy encompassing advisory and educational site-based campaigns as an initial measure to attain the WHO-endorsed vaccination rates. Full article

Financial fraud is still one of the most important threats to the financial industry, causing enormous economic losses and mounting difficulties for conventional fraud detection systems. The systems tend to face challenges in dealing with the rising amount of transactional data, the problem of class imbalance, and the continually changing nature of fraudulent activity. In order to solve these problems, in this research a cloud hybrid framework for detecting fraud using Long Short-Term Memory (LSTM) networks, Autoencoders, and Cat Swarm Optimization (CSO) is suggested. The purpose of the suggested framework is to provide improved detection performance and flexibility on a benchmark financial dataset, with a design intended to support scalability in real-time applications. The framework uses the Credit Card Fraud Detection Dataset from Kaggle, which consists primarily of numerical features, including anonymized variables (V1–V28), along with time and amount. The LSTM networks learn the sequential relationships of transactions, while Autoencoders learn to detect anomalies in the data unsupervised. CSO is used to optimize key hyperparameters of the hybrid model, including the learning rate (0.0001–0.01), batch size (32–128), number of LSTM layers (1–3), number of hidden units per layer (16–128), dropout rate (0.1–0.5), and fusion weights (0–1 for each weight, with the sum constrained to 1) between the LSTM and Autoencoder outputs. In addition, CSO is applied for feature subset selection and threshold tuning to further enhance model performance. Preprocessing is performed on the data, including normalization and feature scaling prior to model training. The suggested framework has a 96.2% accuracy, 94.6% precision, 97.9% recall, 96.2% F1-score, and 0.97 AUC-ROC, showing improved performance compared to CNN-based and LSTM-CNN models under the evaluated conditions. However, since no multiple experiments were conducted to verify the robustness, the results should be interpreted as indicative rather than definitive. The framework exhibits competitive fraud detection performance on the evaluated benchmark dataset, particularly in handling class imbalance. In a simulated environment configured to mimic cloud-like conditions, the framework achieved inference latency between 15 and 30 ms, GPU utilization between 60% and 70%, and a data transfer volume of approximately 1.5 GB per day, suggesting its potential for deployment in cloud-based fraud detection systems. The framework indicates immense potential for cloud deployment, with a robust solution for preventing financial fraud. The proposed framework demonstrates the potential of integrating sequential modeling, anomaly detection, and metaheuristic optimization within a unified and cloud-oriented architecture, providing a more comprehensive approach compared to conventional hybrid models. Full article

Background: Upper-limb impairment is a major contributor to chronic disability after stroke. Conventional recovery protocols frequently suffer from poor adherence, limited accessibility, and insufficient intensity for prolonged rehabilitation. Methods: We performed a systematic analysis of randomized controlled trials (RCTs) and non-randomized designs published between 2019 and 2024, assessing virtual reality (VR) interventions for upper-limb stroke rehabilitation. Participant characteristics, VR intervention details, primary and secondary outcomes, and adherence rates were analyzed in accordance with PRISMA guidelines. The review is registered in PROSPERO (CRD420251150877). We searched PubMed, Embase, Wiley, Scopus, and Cochrane databases. Study quality was assessed using the RoB 2 and ROBINS-I tools. This review received no funding. Results: Forty-one trials met the inclusion criteria. High variability in study methodology, VR devices, intervention protocols, and outcome measures limited direct comparability. Dropout rates were low and were frequently attributed to factors unrelated to the VR intervention. Adverse events were uncommon, supporting the feasibility and safety of VR-based rehabilitation. Conclusions: While VR is a safe and feasible modality, large-scale, multicenter clinical trials with standardized protocols and long-term follow-up are essential to define the role of VR in routine stroke care. Full article

The Individuals with Disabilities Education Act (IDEA) provides special education services to students with disabilities, including autistic students, until age 21. However, the ages at which autistic students exit high school—and the reasons for exit—are not well documented, despite their importance for transition planning. We analyzed U.S. Department of Education Section 618 Part B data for special education students ages 14–21 across five school years (2014–2015 to 2018–2019) to examine exit age and exit category, with comparisons among autistic students, students with intellectual disabilities (IDs), and students with other disabilities. Using publicly reported counts of students exiting at each age, we derived mean exit ages by transforming age-specific count data. In 2019, 71% of autistic students graduated with a diploma, compared with 48% of students with IDs and 72.5% of students with other disabilities. Autistic students had lower dropout rates (6–8%) than students with other disabilities (15–18%). The mean exit age for autistic students was approximately 18 years, with an average graduation age of 17.9 years, indicating that many students exited prior to the end of extended IDEA eligibility in their state. These findings provide descriptive context on when autistic students exit high school relative to IDEA eligibility and underscore the importance of transition planning and coordination with adult service systems, though these factors were not directly examined in the present analysis. Full article

College dropout is a global challenge due to its high prevalence and its consequences for individuals, institutions, and society, particularly in terms of institutional sustainability, inefficient use of public resources, and human capital loss. This issue is especially salient in engineering, where first-year dropout rates remain high. This study examines factors associated with first-year dropout among engineering students at a Chilean public university, framing dropout as a sustainability challenge for higher education systems. The analysis combines administrative records ($n=825$) with survey data on psychosocial variables ($n=417$). Results show that admission to a first-choice program and early performance are strongly associated with persistence, highlighting admission alignment and early university experience as factors contributing to the sustainable use of institutional resources. Despite equivalent academic performance across genders, a marked discrepancy emerged between students’ high self-reported confidence and limited implementation of learning strategies. Cluster analysis identified a clear performance gradient across socio-educational profiles, with students combining high academic capital, low socioeconomic vulnerability, and first-choice admission showing the most favorable outcomes. These findings underscore the relevance of admission preference, trajectories, and socio-educational context for first-year persistence, with implications for institutional sustainability and the consolidation of human capital in engineering education. Full article

Hybrid cognitive behavioral therapy (CBT), combining group and individual sessions, for treating obsessive–compulsive disorder (OCD) has rarely been examined in routine clinical practice. This prospective observational study preliminarily evaluated the feasibility and acceptability of a hybrid CBT program implemented in Japan. The program consisted of one pre-treatment individual session, eight group sessions, and one post-treatment individual session. Feasibility and acceptability were assessed using dropout rates and written questionnaire feedback. Twenty-eight individuals (mean age = 36.1 ± 14.0 years) participated, with two dropouts. Seven participants reported that the program duration was too short, whereas the remaining participants considered it appropriate. Nineteen participants indicated their willingness to participate in a similar program. Open-ended feedback highlighted the importance of group composition and program content. Self-Rating Yale–Brown Obsessive Compulsive Scale scores decreased at Session 8 (estimate = −2.74, p = 0.002) and post-treatment (estimate = −4.16, p < 0.001) according to a linear mixed-effects model. Reductions were also observed in Sheehan Disability Scale, State–Trait Anxiety Inventory, and Clinical Global Impressions Scale scores, whereas Center for Epidemiologic Studies Depression Scale scores showed no significant change. These findings suggest the feasibility and acceptability of the program and may inform future program development. Full article

Utility-scale photovoltaic (PV) plants rely on supervisory control and data acquisition (SCADA) streams for performance verification, yet high-frequency measurements are routinely affected by sensor dropouts, intermittency, and operating-state transitions that bias regression-based capacity estimates. This study evaluates a reproducible SCADA processing workflow for capacity-style reporting and a complementary soiling–clean temperature prediction model using data from a documented October 2022 test window (5 s SCADA aggregated to 1 min). The following three filtering approaches are compared: (i) naïve thresholds (Baseline A), (ii) deterministic stability screening using ramp-rate and rolling-variability constraints (Baseline B), and (iii) an optional residual-based outlier trimming step (Method C). Capacity is estimated via a multivariate regression evaluated on a fixed-size reporting-condition subset (RC197) with day-coverage constraints. All methods achieved high fit quality on RC197 ($R^2\approx 0.99$), with Baseline B improving error and uncertainty over Baseline A (RMSE 2.05 vs. 2.18 MW; $U_{95}$ 0.97% vs. 1.03%) while preserving day coverage; Method C yielded the lowest in-sample RMSE (1.89 MW) but reduced day coverage. For temperature prediction, a baseline-plus-residual learning formulation substantially improved leave-one-day-out performance, reducing MAE/RMSE from 2.99/3.76 °C to 1.43/1.80 °C and increasing $R^2$ from 0.60 to 0.91. The results highlight trade-offs between fit tightness and representativeness in capacity-style filtering and demonstrate residual learning is an effective approach for SCADA-based thermal characterization. Full article

We present the first secure multiplicative aggregation protocol as a variant of secure aggregation. In this case, a server can compute the component-wise product of the input vectors of users while handling the possible dropout of users during protocol execution. Using pairwise masks, threshold secret sharing and the secure aggregation protocol itself, our construction is correct and secure against semi-honest adversaries. We also consider secure aggregation protocols for the case in which fixed users can reuse their private keys to do aggregation many times, and we propose key reusable secure aggregation protocols. Our protocols have an overhead polynomial in the number of users. We conduct a comprehensive evaluation of our proposed protocols. For multiplicative aggregation protocol, experiments varying the number of users (K) from 50 to 300 (with fixed input size $X_u=100$ KB) demonstrate that user computation scales monotonically with K and is largely insensitive to dropout rates. In contrast, server computation is highly dropout-sensitive and exhibits a steeper growth rate with respect to K. When varying the input size (10–250 KB) with a fixed K, both user and server communication overheads increase linearly, while server computation remains the primary bottleneck affected by dropouts. We compare reusable and non-reusable secure aggregation protocol over repeated interactions $q\in \{1,\dots ,10\}$ at $X_u=100$ KB and $K=100$, showing that reusing Round 1 reduces the cumulative user computation time by about 2.5 times and reduces the cumulative server computation overhead by about 1.2 times at $q=10$ while leaving the server communication overhead nearly unchanged, which indicates that the overall communication overhead is dominated by the non-reused rounds. Full article

Background: Responders to the September 11, 2001, WTC attacks experience high rates of PTSD, and existing treatments often lead to high dropout and low care use. Objectives: This randomized, double-blind, sham-controlled trial assesses the feasibility and acceptability of transcutaneous auricular vagus nerve stimulation (taVNS) as a potential PTSD treatment for 9/11 responders. Methods: A total of 32 WTC responders aged 18+ with PTSD, recruited via the World Trade Center Health Program, participated; those with current psychosis, unstable medical conditions, or recent trial involvement were excluded. Participants were randomly assigned to taVNS or sham groups and asked to use the device for 15 min daily for 8 weeks, with staff and participants blinded. Primary outcomes included recruitment, adherence, retention, and feedback. Secondary outcomes examined changes in depression (PHQ-9), anxiety (GAD-7), and sleep (PSQI). Data were analyzed with mixed-effects models focusing on PTSD and mental health symptoms. Results: The taVNS group showed modest PTSD improvement, with a 10-point CAPS-5 reduction in 40% of stimulation participants versus 28.5% sham; no significant differences in self-reported symptoms were found. Discussion: Daily taVNS over eight weeks is feasible and acceptable, warranting larger studies to detect differences and identify subgroups with greater benefit. Trial registration: “taVNS to Reduce PTSD Symptoms in WTC Responders” (NCT05212714); registered 9 September 2021. Full article

Accurate and robust brain tumor segmentation remains a critical challenge in medical image analysis due to high inter-patient variability, complex tumor morphology, and modality-specific noise in MRI scans. This study proposes PSO-GA-U-Net, a novel hybrid deep learning framework that integrates Particle Swarm Optimization (PSO) and Genetic Algorithms (GAs) to optimize the U-Net architecture, enhancing segmentation performance and generalization. PSO dynamically tunes the learning rate to accommodate modality-specific variations, while the GA adaptively regulates dropout to improve feature diversity and reduce overfitting. The model was evaluated on three benchmark datasets—FBTS, BraTS 2021, and BraTS 2018—using five-fold cross-validation. PSO-GA-U-Net achieves Dice Similarity Coefficients (DSC) of 0.9587, 0.9406, and 0.9480 and Jaccard Index (JI) scores of 0.9209, 0.8881, and 0.9024, respectively, consistently outperforming state-of-the-art models in both overlap accuracy and boundary delineation. Statistical tests confirm that these improvements are significant across folds ($p<0.05$). Visual heatmaps further illustrate the model’s ability to preserve structural integrity across tumor types and modalities. These results indicate that metaheuristic-guided deep learning offers a promising and clinically applicable solution for automatic tumor segmentation in radiological workflows. Full article

Background/Objectives: People with mobility limitations face disproportionately high rates of chronic health conditions and demonstrate lower adherence to wellness interventions. Digital programs such as MENTOR offer accessible alternatives but often face high rates of attrition. This study applied machine learning (ML) methods to predict adherence to the eight-week MENTOR telewellness program and identify key predictors of participant attendance. Methods: Data were drawn from 1218 adults enrolled in MENTOR (2023–2024). Adherence was defined as the percentage of 40 sessions attended. Baseline demographic, socioeconomic, psychosocial, mindfulness, resilience, health status, and physical activity variables were included as predictors. Following preprocessing and imputation, 13 ML regression models were trained using an 80/20 train–test split. The best-performing model was identified using mean absolute error (MAE), followed by feature selection and SHAP interpretability analyses. Pairwise synergy analysis quantified interactions between top predictors. Results: Model performance was modest overall. Bayesian ridge regression achieved the best performance (MAE 20.98; RMSE 25.26; R² = 0.12). SHAP analyses revealed that education, race, emotional support, Area Deprivation Index, household size, mindfulness, life satisfaction, and disability onset were the strongest predictors of adherence. Higher emotional support, mindfulness, and life satisfaction were associated with greater adherence, while socioeconomic disadvantage predicted lower adherence. Synergy analyses showed the strongest predictive interactions between low education and psychosocial resources (emotional support and life satisfaction). Conclusions: Baseline characteristics alone modestly predicted adherence to a digital wellness program. However, psychosocial and socioeconomic factors emerged as meaningful predictors, underscoring the need for personalized support strategies to reduce dropout among participants with mobility limitations. Full article

Educational success is a critical component of societal development, yet increasing student dropout rates present ongoing challenges. While supervised learning models are commonly used for dropout prediction, they rely on manually labeled data, a process that is time-consuming and dependent on expert annotation. Unsupervised learning models, clustering approaches, have been explored as an alternative; however, existing methods typically group students based on activity patterns without generating binary outcome labels such as dropout or success. Furthermore, their effectiveness often depends heavily on the quality of the selected features, and most current solutions utilize only limited or pre-structured subsets of institutional data. This paper addresses these challenges and proposes EHCFE (Enhanced Hierarchical Clustering with Feature Engineering), to automatically generate binary labels from unlabeled educational datasets. EHCFE applies feature engineering by generating new features from the top-ranked features identified during feature selection while retaining the original feature set, thereby improving the quality of the labeling outcomes. The approach is evaluated on three datasets and compared with current and state-of-the-art models using several evaluation metrics, including F₁ score, area under the receiver operating characteristic curve (AUC), and silhouette coefficient. Experimental results show that EHCFE achieves the highest F₁ score (0.709 and 0.28) and AUC values (0.766 and 0.81) on two datasets. A ranking analysis across six evaluation metrics demonstrates that EHCFE outperforms existing models, achieving the highest average ranks of 1.50 and 1.83 on two datasets and a competitive rank of 1.92 on the third. Full article

A tool-health digital twin for CNC predictive maintenance is developed and operationalised as a fusion-and-state-estimation core that produces a latent tool-health trajectory (wear level and wear-rate dynamics) from multi-rate sensor streams for diagnosis and remaining useful life (RUL) forecasting under strict edge latency constraints. The scope is tool-health–informed maintenance decisions (condition-based tool replacement/scheduling), rather than a comprehensive maintenance twin for all CNC subsystems. Multi-rate vibration, spindle-current, and temperature signals are synchronized and windowed, and a linear state-space model with Kalman filtering and innovation-guided adaptive noise estimation stabilizes the latent health state across operating-regime changes. The fused state is then used by compact sequence learners, an LSTM for edge feasibility, and a compact Transformer as a higher-accuracy comparison, to output fault categories and RUL estimates. Predictive uncertainty is quantified via a Monte Carlo dropout and linked to reliability-aware actions through a simple alarm/defer/schedule policy, while SHAP provides feature-level interpretability. On a CNC testbed, fusion improves fault F1 from 0.811 to 0.892 and PR-AUC from 0.867 to 0.918 while reducing RUL RMSE from 10.4 to 8.1 cycles; the compact Transformer reaches 0.903 F1 and 7.9-cycle RMSE at higher inference time. The end-to-end pipeline remains within a ≤100 ms breakdown, maintains in-band innovation statistics, supports rehearsal-based updates under drift, and is additionally evaluated on external tool-wear and turbofan datasets. Full article

In the context of smart manufacturing, with the widespread deployment of Industrial Internet of Things (IoT) devices, a large number of computation tasks that are highly sensitive to latency and have strict deadlines have emerged, requiring real-time processing. Effectively offloading tasks to address the issues of increased latency and task dropouts caused by dynamic changes in edge node load has become a key challenge in the cloud–edge–end collaborative environment of smart manufacturing. To tackle the complex issues of unknown edge node loads and dynamic system state changes, this paper proposes a distributed algorithm based on deep reinforcement learning, utilizing convolutional neural networks (CNN) and the Informer architecture. The proposed algorithm leverages CNN to extract local features of edge node loads while utilizing Informer’s self-attention mechanism to capture long-term load variation trends, thereby effectively handling the uncertainty and dynamics inherent in node loads. Furthermore, by integrating the Dueling Deep Q-Network (DQN) and Double DQN techniques, the algorithm achieves a precise approximation of the state–action value function, further enhancing its capability to perceive system temporal characteristics and adapt to heterogeneous tasks. Each mobile device can independently make task offloading decisions and scheduling strategies based on its observations, enabling dynamic task allocation and optimization of execution order. Simulation results show that, compared to various existing algorithms, the proposed method reduces task dropout rates by 82.3–94% and average latency by 28–39.2%. Experimental results validate the significant advantages of this method in intelligent manufacturing scenarios with high load and latency-sensitive tasks. Full article