Search Results (5,601)

Background/Objectives: Photoimmunotherapy for head and neck cancer (HN-PIT) is an emerging treatment for unresectable locally advanced or recurrent head and neck cancer. However, real-world data (RWD) are limited. This study examined the safety and effectiveness of HN-PIT. Methods: This multicenter, retrospective cohort study included 40 patients with unresectable locally advanced or recurrent head and neck cancers who underwent HN-PIT from January 2021 to August 2024. The primary endpoint was time to treatment failure (TTF). Secondary endpoints included the objective response rate (ORR), overall survival (OS), progression-free survival (PFS), and adverse events (AEs). Results: The median TTF and 1-year treatment failure rate were 6.0 months and 23.2%, respectively. Moreover, the ORR, disease control rate, median OS, and median PFS were 75.0% (95% confidence interval [CI]: 60.0–86.0%), 95.0% (95% CI: 83.5–99.0%), 26.9 months, and 6.2 months, respectively. The incidence of grade ≥3 AEs was 17.5% (95% CI: 7.1–29.1%). Pain was the most common AE, occurring in 37 patients (92.5%), with grade III pain reported in 5 (12.5%). Mucositis occurred in 32 patients (80.0%), with grade III mucositis reported in 3 (7.5%). Hemorrhages occurred in 31 patients (77.5%), with no grade ≥III hemorrhages reported. Two patients experienced sepsis (5.0%; grades IV and V). Seventeen patients (42.5%) had laryngeal edema, with grade IV edema reported in four (10.0%). Conclusions: Our RWD shows that HN-PIT is effective with an acceptable safety profile. TTF may serve as an endpoint reflecting this treatment’s characteristics. This study provides important basic data for the development of future treatment strategies. Full article

Dam displacement serves as a critical visual indicator for assessing structural integrity and stability in dam engineering. Data-driven displacement forecasting has become essential for modern dam safety monitoring systems, though conventional approaches—including statistical models and basic machine learning techniques—often fail to capture comprehensive feature representations from multivariate environmental influences. To address these challenges, a bidirectional gated recurrent unit (BiGRU)-enhanced neural network is developed, incorporating sliding window mechanisms to model time-dependent hysteresis characteristics. The BiGRU’s architecture systematically integrates historical temporal patterns through overlapping window segmentation, enabling dual-directional sequence processing via forward–backward gate structures. Validated with four instrumented measurement points from a major concrete gravity dam, the proposed model exhibits significantly better performance against three widely used recurrent neural network benchmarks in displacement prediction tasks. These results confirm the model’s capability to deliver high-fidelity displacement forecasts with operational stability, establishing a robust framework for infrastructure health monitoring applications. Full article

Water quality deterioration poses a critical threat to ecological security and sustainable development, particularly in rapidly urbanizing regions. To enable proactive environmental management, this study develops a novel hybrid deep learning model, the NGO-CNN-GRU, for high-precision time-series water quality prediction in the Xijiang River Basin, China. The model integrates a Convolutional Neural Network (CNN) for spatial feature extraction and a Gated Recurrent Unit (GRU) for temporal dependency modeling, with hyperparameters optimized via the Northern Goshawk Optimization (NGO) algorithm. Using historical water quality (pH, DO, CODMn, NH3-N, TP, TN) and meteorological data (precipitation, temperature, humidity) from 11 monitoring stations, the model achieved exceptional performance: test set R² > 0.986, MAE < 0.015, and RMSE < 0.018 for total nitrogen prediction (Xiaodong Station case study). Across all stations and indicators, it consistently outperformed baseline models (GRU, CNN-GRU), with average R² improvements of 12.3% and RMSE reductions up to 90% for NH3-N predictions. Spatiotemporal analysis further revealed significant pollution gradients correlated with anthropogenic activities in the Pearl River Delta. This work provides a robust tool for real-time water quality early warning systems and supports evidence-based river basin management. Full article

Intensified by climate change and anthropogenic activities, flood disasters necessitate rapid and accurate mapping for effective disaster management. This study develops an integrated framework leveraging synthetic aperture radar (SAR) and cloud computing to enhance flood monitoring, with a focus on a 2024 extreme rainfall event in Liaoning Province, China. Utilizing the Google Earth Engine (GEE) platform, we combine three complementary techniques: (1) Otsu automatic thresholding, for efficient extraction of surface water extent from Sentinel-1 GRD time series (154 scenes, January–October 2024), achieving processing times under 2 min with >85% open-water accuracy; (2) random forest (RF) classification, integrating multi-source features (SAR backscatter, terrain parameters from 30 m SRTM DEM, NDVI phenology) to distinguish permanent water bodies, flooded farmland, and urban areas, attaining an overall accuracy of 92.7%; and (3) Fuzzy C-Means (FCM) clustering, incorporating backscatter ratio and topographic constraints to resolve transitional “mixed-pixel” ambiguities in flood boundaries. The RF-FCM synergy effectively mapped submerged agricultural land and urban spill zones, while the Otsu-derived flood frequency highlighted high-risk corridors (recurrence > 10%) along the riverine zones and reservoir. This multi-algorithm approach provides a scalable, high-resolution (10 m) solution for near-real-time flood assessment, supporting emergency response and sustainable water resource management in affected basins. Full article

Background: Invasive BC patients are at risk of loco-regional recurrence, distant MTS, and the development of second primary tumors. SPMs comprise the sixth most common group of malignancies. Material and methods: The records of 125 consecutive patients with primary invasive TCC of the bladder seen in the Oncology Department of Soroka University Medical Center were reviewed between January 2016 and December 2023. We recorded demographic details, the type of primary treatment, tumor site, time to diagnosis of MTS, and occurrence of SPMs. Results: The primary treatments included RC in 58 patients (median age 66 years, range 43–86), PC in 9 patients (median age 64 years, range 22–73), and XRT in 23 patients (median age 74 years, range 22–87). Five patients from the PC group were also treated by XRT. A total of 90 (72%) patients developed MTS or SPMs, with 66 of these developing MTS and 24 developing SPMs. The median age was 70 years (range 22–87). The most frequent site of MTS was in the pelvic LNs (34 patients), followed by bone (18 patients), liver (8 patients), and lung (6 patients), with 4 patients developing synchronous MTS in the pelvic LNs and liver. The median time from diagnosis to MTS was 14.3 months. The distribution of MTS varied according to primary treatment. After RC, 17 patients developed LN MTS, 7 liver, 6 bone, and 3 lung MTS. The average times for developing MTS were as follows: LNs, 14.8 months, liver, 59.7 months, bone, 6.8 months, and lung, 16 months. Following XRT, LN MTS developed in 17 patients: 12 bone, 3 lung, and 1 liver. The most frequent SPMs were prostate cancer with 11 patients and lung cancer with 6 patients, with the median time from TCC diagnosis of 54 months. Conclusion: A regular extended follow-up for invasive BC patients is vital to ensure the early detection of frequently occurring MTS and SPMs. Through the early diagnosis of local recurrences, MTS, and SPMs, treatment results and patient prognosis can be significantly improved. Full article

Background: Hamstring strain injuries (HSIs) are among the most common and recurrent injuries in sports involving high-speed running. While eccentric training has demonstrated efficacy in reducing HSI risk, the role of sprint training as a preventive strategy remains underexplored and often misinterpreted as solely a risk factor. Methods: This review aimed to systematically map the available evidence on the role of sprint training in hamstring injury prevention, identifying mechanisms, outcomes, and potential synergies with other strategies. This scoping review was conducted following the Joanna Briggs Institute’s methodology and reported in accordance with PRISMA-ScR guidelines. Seven databases (PubMed, Scopus, Web of Science, Cochrane CENTRAL, SPORTDiscus, CINAHL, and PEDro) were searched up to October 2024. Studies were included if they involved adult athletes and examined the effects of sprint training, ≥80–90% maximal sprint speed (MSS), on hamstring injury prevention, muscle architecture, or functional outcomes. All databases were searched from inception to 15 October 2024, and the screening and data-charting process was completed on 30 April 2025. Results: Twelve studies met the inclusion criteria. Sprint exposure, when combined with eccentric strengthening and biomechanical optimisation, led to injury reductions ranging from 56% to 94%. Eccentric interventions produced fascicle length increases of up to 20% and strength gains of 15–20%. Improvements in sprint technique and neuromuscular control were also reported. Biomechanical risk factors, including pelvic tilt and hip extension deficits, were linked to increased HSI risk. The most common eccentric protocols included Nordic Hamstring Exercises (NHE), Razor Curls, and hip-dominant exercises, typically performed 1–2 times per week for 4 to 8 weeks. Conclusions: High-speed sprint training, when properly programmed and integrated into comprehensive preventive strategies, may enhance tissue resilience and reduce HSI risk. Combining sprint exposure with eccentric strengthening and technical coaching appears to be more effective than isolated interventions alone. Practically, these results support the systematic inclusion of progressive high-intensity sprint exposure in routine hamstring-injury-prevention programmes for field-sport athletes. Full article

Background: Hepatocellular carcinoma is one of the leading causes of cancer-related deaths worldwide. Its high recurrence rate and limited treatment options underscore the urgent need for the development of new and highly effective drugs. Methods: This study systematically explores the molecular mechanism of cinnamic acid against hepatocellular carcinoma through integrated machine learning prediction, network pharmacological analysis and in vitro experimental verification. Results: The prediction of anti-tumor activity based on the random forest model showed that cinnamic acid has significant anti-tumor potential (probability = 0.69). Network pharmacology screened 185 intersection targets of cinnamic acid and liver cancer, of which 39 core targets (such as PIK3R1, AKT1, MAPK1) were identified as key regulatory hubs through protein interaction network and topological analysis. Functional enrichment analysis showed that these targets were mainly enriched in the PI3K/AKT signaling pathway (p = 2.1 × 10⁻¹²), the cancer pathway (p = 3.8 × 10⁻¹⁰), and apoptosis-related biological processes. Molecular docking validation revealed that the binding energies of cinnamic acid with the 19 core targets were all below −5 kcal/mol, a threshold indicating strong binding affinity in molecular docking. The binding modes to PIK3R1 (−5.4 kcal/mol) and AKT1 (−5.1 kcal/mol) stabilized through hydrogen bonding. In vitro, cinnamic acid dose-dependently inhibited Hep3B proliferation/migration, induced apoptosis, downregulated PI3K, p-AKT, and Bcl-2, and upregulated Bax and Caspase-3/8. Conclusions: This study systematically reveals, for the first time, that the multi-target mechanism of cinnamic acid exerts anti-hepatic cancer effects by targeting the PI3K/AKT signaling pathway, supporting its potential as a natural anti-tumor drug. Full article

Sign language is a three-dimensional (3D) visual language that conveys meaning through hand positions, shapes, and movements. Traditional sign language education methods, such as textbooks and videos, often fail to capture the spatial characteristics of sign language, leading to limitations in learning accuracy and comprehension. To address this, we propose a 3D Korean Sign Language Learning System that leverages pseudo-hologram technology and hand gesture recognition using Leap Motion sensors. The proposed system provides learners with an immersive 3D learning experience by visualizing sign language gestures through pseudo-holographic displays. A Recurrent Neural Network (RNN) model, combined with Diffusion Convolutional Recurrent Neural Networks (DCRNNs) and ProbSparse Attention mechanisms, is used to recognize hand gestures from both hands in real-time. The system is implemented using a server–client architecture to ensure scalability and flexibility, allowing efficient updates to the gesture recognition model without modifying the client application. Experimental results show that the system enhances learners’ ability to accurately perform and comprehend sign language gestures. Additionally, a usability study demonstrated that 3D visualization significantly improves learning motivation and user engagement compared to traditional 2D learning methods. Full article

Ontology technology addresses data heterogeneity challenges in Internet of Everything (IoE) systems enabled by Cyber Twin and 6G, yet the subjective nature of ontology engineering often leads to differing definitions of the same concept across ontologies, resulting in ontology heterogeneity. To solve this problem, this study introduces a hybrid ontology matching method that integrates a Recurrent Neural Network (RNN) with syntax-based analysis. The method first extracts representative entities by leveraging in-degree and out-degree information from ontological tree structures, which reduces training noise and improves model generalization. Next, a matching framework combining RNN and N-gram is designed: the RNN captures medium-distance dependencies and complex sequential patterns, supporting the dynamic optimization of embedding parameters and semantic feature extraction; the N-gram module further captures local information and relationships between adjacent characters, improving the coverage of matched entities. The experiments were conducted on the OAEI benchmark dataset, where the proposed method was compared with representative baseline methods from OAEI as well as a Transformer-based method. The results demonstrate that the proposed method achieved an 18.18% improvement in F-measure over the best-performing baseline. This improvement was statistically significant, as validated by the Friedman and Holm tests. Moreover, the proposed method achieves the shortest runtime among all the compared methods. Compared to other RNN-based hybrid frameworks that adopt classical structure-based and semantics-based similarity measures, the proposed method further improved the F-measure by 18.46%. Furthermore, a comparison of time and space complexity with the standalone RNN model and its variants demonstrated that the proposed method achieved high performance while maintaining favorable computational efficiency. These findings confirm the effectiveness and efficiency of the method in addressing ontology heterogeneity in complex IoE environments. Full article

Background: Paroxysmal atrial fibrillation (PAF) is a common arrhythmia often treated with catheter ablation, particularly pulmonary vein isolation (PVI). However, recurrence remains frequent and is often linked to unrecognized structural and functional remodeling of the left atrium. Methods: We introduce the Echocardiographic Atrial Strain and conduction Evaluation (EASE) score as a theoretical, noninvasive model to stratify recurrence risk in patients undergoing catheter ablation for PAF. The score is based on the hypothesis that integrated echocardiographic parameters can reflect the extent of atrial remodeling relevant to ablation outcomes. Results: The EASE score combines six echocardiographic metrics—left atrial reservoir strain (LASr), atrial conduction time (PA-TDI), left atrial volume index (LAVI), stiffness index (E/e′/LASr), E/e′ ratio, and contractile strain (LASct)—each representing structural, electrical, or mechanical remodeling. The total score ranges from 0 to 12, stratifying patients into low, intermediate, and high-risk categories for arrhythmia recurrence. Preliminary retrospective data suggest a significant association between higher EASE scores and increased recurrence rates following ablation. Conclusions: The EASE score offers a biologically plausible, multidimensional framework for noninvasive risk prediction in PAF ablation. Prospective studies are warranted to validate its clinical utility and refine its structure. Full article

Backgrounds. Lateral neck dissection (LND) is standard for thyroid cancer patients with neck metastases, mostly performed at the same time as total thyroidectomy (cLND). We introduced a new delayed LND (dLND), 4 weeks after thyroidectomy to reduce surgical morbidity. This study aims to compare the oncologic/complication outcomes between the two strategies, based on a large retrospective cohort of patients. Methods. Between 1996 and 2024, 215 patients were treated with total thyroidectomy, central neck dissection (CND) and LND, and grouped by surgical strategy (cLND vs. dLND); survival/complication outcomes were analyzed and compared between the two groups. Results. The overall and disease-free survival were comparable between groups. Age, extracapsular extension, and nodal burden predicted recurrence. dLND was associated with a significantly lower risk of vocal fold palsy. Extranodal extension (ECE) strongly predicted nerve injury. Conclusions. dLND offers similar oncologic outcomes to cLND, with reduced risk of vocal fold palsy. A staged approach enhances nerve preservation and might be considered in treatment planning. Full article

Objective: To evaluate the perioperative outcomes, functional impact, and oncologic efficacy of off-clamp robotic-assisted partial nephrectomy (RAPN) in patients with renal masses across multiple high-volume centers. Materials and Methods: We conducted a retrospective multicenter study including 563 patients (group 1) who underwent clampless RAPN between January 2018 and December 2024. Patients with solitary kidneys, tumors >7 cm, or prior renal surgery were excluded. The standardized surgical technique involved tumor resection without clamping of the renal artery, followed by the use of hemostatic agents and standard/selective suturing of the resection bed on demand. Patients in group 1 were compared to 244 consecutive patients treated in the same centres and treated with RAPN with an on-clamp procedure (group 2). Primary outcomes included operative time, blood loss, and complications, while secondary outcomes assessed renal function preservation and oncologic control at an at least 12-month follow-up. Results: The median operative time was 118 min (IQR: 100–140 min), and median estimated blood loss was 150 mL (range: 50–400 mL). The overall complication rate was 9.2%, with most classified as Clavien–Dindo Grade I–II. No intraoperative conversions to open surgery were recorded. Renal function was well preserved, with a median estimated glomerular filtration rate (eGFR) decline of 4.1% at three months (p > 0.05), and no cases of acute kidney injury. Oncologic outcomes were favorable, with a positive surgical margin rate (PSM) of 2.4% and two cases of tumor recurrences (0.36%) documented at a 12-month follow-up. Conclusions: The off-clamp RAPN is a safe and effective nephron-sparing approach, offering significant renal function preservation while maintaining oncologic efficacy. This technique minimizes ischemia–reperfusion injury and post-surgical fibrosis, providing a viable alternative to on-clamp RAPN. Further prospective trials are warranted to confirm long-term benefits and refine patient selection criteria. Full article

This paper introduces a speed-sensorless dual-field-oriented control (DFOC) strategy for induction motors (IMs). DFOC combines the advantages or rotor- and stator-field orientation to significantly reduce the parameter sensitivity of the control regarding the generation of the converter control variable. A simplified structure is also proposed, using only two regulators for the flux and speed control, eliminating the two current regulators. Related to sensorless control, the classical adaptation mechanism within an MRAS (model reference adaptive system) observer is replaced with artificial intelligence (AI)-based approaches. Specifically, artificial neural networks (ANNs) and recurrent neural networks (RNNs) are employed for rotor speed estimation. They offer significant advantages in managing complex and nonlinear systems, providing enhanced flexibility and adaptability compared to traditional MRAS methods. The effectiveness of the proposed sensorless control scheme is validated through both simulation and real-time implementation. The paper focuses on the ANN and RNN architectures, as deep learning models, in terms of the reliability and accuracy of rotor speed estimation under various operating conditions. Full article

Background/Objectives: Acute exacerbation of interstitial lung disease (AE-ILD) is a life-threatening complication in lung cancer patients with pre-existing ILD. Anatomical resection is recognized as a significant risk factor for AE-ILD. We investigated the safety and feasibility of wedge resection in lung cancer patients with ILD. Methods: This retrospective study analyzed clinical stage IA–IIIA primary lung cancer patients with ILD, as recorded in the Shizuoka Registry across eight institutions from January 2019 to May 2023. Patients were categorized into a wedge resection group (WG) and an anatomical resection group (AG), which included segmentectomy, lobectomy, and bilobectomy. Perioperative outcomes were compared between the groups. Results: The WG comprised 36 patients, while the AG included 81. The WG had significantly older patients (77 vs. 72 years, p < 0.01) and smaller tumors (18 vs. 24 mm, p < 0.01). Wedge resection was associated with shorter operative time (100 vs. 205 min, p < 0.01) and less blood loss (5 vs. 30 mL, p = 0.02). The incidence of postoperative complications did not differ significantly (p = 0.84). AE-ILD occurred in three patients (8%) in the WG and four patients (4%) in the AG. Perioperative mortality was 0% in the WG and 2% in the AG; both deaths were due to AE-ILD. Marginal recurrence was observed in four patients (11%) in the WG. Conclusions: Although AE-ILD incidence was higher, no deaths due to IP-AE were observed in the WG. While wedge resection cannot completely prevent postoperative AE-ILD, it may reduce perioperative mortality in lung cancer patients with ILD. Full article

The precise forecasting of renewable energy production and usage is essential for the stability, efficiency, and sustainability of contemporary power systems. This requirement is especially urgent in South Africa, a nation currently grappling with considerable energy issues, such as recurrent load shedding, outdated coal-fired power plants, and an increasing electricity demand. As the country moves towards a more renewable-focused energy portfolio, the capacity to anticipate future energy requirements is crucial for effective planning, operational stability, and grid resilience. This study introduces a hybrid approach that combines deep learning and machine learning techniques, specifically integrating long short-term memory (LSTM) neural networks with extreme gradient boosting (XGBoost) to provide more accurate and detailed forecasts of energy demand. LSTM networks are particularly effective in capturing long-term temporal dependencies in sequential data, such as patterns of energy usage. At the same time, XGBoost delivers high-performance gradient-boosted decision trees that can manage non-linear relationships and noise present in extensive datasets. The proposed hybrid LSTM-XGBoost model was trained and assessed using high-resolution data on energy consumption and weather conditions gathered from a coastal municipality in KwaZulu-Natal, South Africa, a country that exemplifies the convergence of renewable energy potential and challenges related to energy reliability. The preprocessing steps, including normalization, feature selection, and sequence modeling, were implemented to enhance the input data for both models. The performance of the model was thoroughly evaluated using standard statistical metrics, specifically the mean absolute error (MAE), the root mean squared error (RMSE), and the coefficient of determination (R²). The hybrid model achieved an MAE of merely 192.59 kWh and an R² of approximately 0.71, significantly surpassing the performance of the individual LSTM and XGBoost models. These findings highlight the enhanced predictive capabilities of the hybrid model in capturing both temporal trends and feature interactions in energy consumption behavior. Full article