Search Results (4,830)

Head and neck cancer (HNC) is highly heterogeneous and difficult to treat, underscoring the need for rapid, patient-specific models. Standard three-dimensional (3D) cultures often lose stromal partners that influence therapy response. We developed a patient-derived system maintaining tumor cells, cancer-associated fibroblasts (CAFs), and cells undergoing partial epithelial–mesenchymal transition (pEMT) for drug sensitivity testing. Biopsies from four HNC patients were enzymatically dissociated. CAFs were directly cultured, and their conditioned medium (CAF-CM) was collected. Cryopreserved primary tumor cell suspensions were later revived, screened in five different growth media under 2D conditions, and the most heterogeneous cultures were re-embedded in 3D hydrogels with varied gel mixtures, media, and seeding geometries. Tumoroid morphology was quantified using a perimeter-based complexity index. Viability after treatment with cisplatin or Notch modulators (RIN-1, recombination signal-binding protein for immunoglobulin κ J region (RBPJ) inhibitor; FLI-06, inhibitor) was assessed by live imaging and the water-soluble tetrazolium-8 (WST-8) assay. Endothelial Cell Growth Medium 2 (ECM-2) medium alone produced compact CAF-free spheroids, whereas ECM-2 supplemented with CAF-CM generated invasive aggregates that deposited endogenous matrix. Matrigel with this medium and single-point seeding gave the highest complexity scores. Two of the three patient tumoroids were cisplatin-sensitive, and all showed significant growth inhibition with the FLI-06 Notch inhibitor, while the RBPJ inhibitor RIN-1 induced minimal change. The optimized scaffold retains tumor–stroma crosstalk and provides patient-specific drug response data within days after operation, supporting personalized treatment selection in HNC. Full article

Electronic-waste (e-waste) management is a key challenge in engineering smart cities due to its rapid accumulation, complex composition, sparse data availability, and significant environmental and economic impacts. This study employs a bespoke machine learning infrastructure on an Indian e-waste supply chain network (SCN) focusing on the three pillars of sustainability—environmental, economic, and social. The economic resilience of the SCN is investigated against external perturbations, like market fluctuations or policy changes, by analyzing six stochastically perturbed modules, generated from the optimal point of the original dataset using Monte Carlo Simulation (MCS). In the process, MCS is demonstrated as a powerful technique to deal with sparse statistics in SCN modeling. The perturbed model is then analyzed to uncover “hidden” non-linear relationships between key variables and their sensitivity in dictating economic arbitrage. Two complementary ensemble-based approaches have been used—Feedforward Neural Network (FNN) model and Random Forest (RF) model. While FNN excels in regressing the model performance against the industry-specified target, RF is better in dealing with feature engineering and dimensional reduction, thus identifying the most influential variables. Our results demonstrate that the FNN model is a superior predictor of arbitrage conditions compared to the RF model. The tangible deliverable is a data-driven toolkit for smart engineering solutions to ensure sustainable e-waste management. Full article

In recent years, the demand for both high accuracy and real-time performance in 3D object detection has increased alongside the advancement of autonomous driving technology. While multimodal methods that integrate LiDAR and camera data have demonstrated high accuracy, these methods often have high computational costs and latency. To address these issues, we propose an efficient 3D object detection network that integrates three key components: a DepthWise Lightweight Encoder (DWLE) module for efficient feature extraction, an Efficient LiDAR Image Fusion (ELIF) module that combines channel attention with cross-modal feature interaction, and a Mixture of CNN and Point Transformer (MCPT) module for capturing rich spatial contextual information. Experimental results on the KITTI dataset demonstrate that our proposed method outperforms existing approaches by achieving approximately 0.6% higher 3D mAP, 7.6% faster inference speed, and 17.0% fewer parameters. These results highlight the effectiveness of our approach in balancing accuracy, speed, and model size, making it a promising solution for real-time applications in autonomous driving. Full article

Sleep staging is a crucial indicator for assessing sleep quality, which contributes to sleep monitoring and the diagnosis of sleep disorders. Although existing sleep staging methods achieve high classification performance, two major challenges remain: (1) the ability to effectively extract salient features from multi-channel sleep data remains limited; (2) excessive model parameters hinder efficiency improvements. To address these challenges, this work proposes a lightweight multi-channel sleep staging network (LMCSleepNet). LMCSleepNet is composed of four modules. The first module enhances frequency domain features through continuous wavelet transform. The second module extracts time–frequency features using multi-scale convolutions. The third module optimizes ResNet18 with depthwise separable convolutions to reduce parameters. The fourth module improves spatial correlation using the Convolutional Block Attention Module (CBAM). On the public datasets SleepEDF-20, SleepEDF-78, and LMCSleepNet, respectively, LMCSleepNet achieved classification accuracies of 88.2% (κ = 0.84, MF1 = 82.4%) and 84.1% (κ = 0.77, MF1 = 77.7%), while reducing model parameters to 1.49 M. Furthermore, experiments validated the influence of temporal sampling points in wavelet time–frequency maps on sleep classification performance (accuracy, Cohen’s kappa, and macro-average F1-score) and the influence of multi-scale dilated convolution module fusion methods on classification performance. LMCSleepNet is an efficient lightweight model for extracting and integrating multimodal features from multichannel Polysomnography (PSG) data, which facilitates its application in resource-constrained scenarios. Full article

Electrocardiogram (ECG) waveform delineation is fundamental to cardiac disease diagnosis. This task requires precise localization of key fiducial points, specifically the onset, peak, and offset positions of P-waves, QRS complexes, and T-waves. Current methods exhibit significant performance degradation in noisy clinical environments (baseline drift, electromyographic interference, powerline interference, etc.), compromising diagnostic reliability. To address this limitation, we introduce ECG-SCFNet: a novel dual-stream architecture employing selective context fusion. Our framework is further enhanced by a consistency training paradigm, enabling it to maintain robust waveform delineation accuracy under challenging noise conditions.The network employs a dual-stream architecture: (1) A temporal stream captures dynamic rhythmic features through sequential multi-branch convolution and temporal attention mechanisms; (2) A morphology stream combines parallel multi-scale convolution with feature pyramid integration to extract multi-scale waveform structural features through morphological attention; (3) The Selective Context Fusion (SCF) module adaptively integrates features from the temporal and morphology streams using a dual attention mechanism, which operates across both channel and spatial dimensions to selectively emphasize informative features from each stream, thereby enhancing the representation learning for accurate ECG segmentation. On the LUDB and QT datasets, ECG-SCFNet achieves high performance, with F1-scores of 97.83% and 97.80%, respectively. Crucially, it maintains robust performance under challenging noise conditions on these datasets, with 88.49% and 86.25% F1-scores, showing significantly improved noise robustness compared to other methods and demonstrating exceptional robustness and precise boundary localization for clinical ECG analysis. Full article

This study aimed to evaluate how different insertion angles of titanium orthodontic miniscrews (30°, 45°, and 90°) influence stress distribution and displacement in surrounding alveolar bone using three-dimensional finite element analysis (FEA), with a focus on biomechanical outcomes at the titanium–bone interface. The 90° insertion angle generated the highest stress in cortical bone (58.2 MPa) but the lowest displacement (0.023 mm), while the 30° angle produced lower stress (36.4 MPa) but greater displacement (0.052 mm). The 45° angle represented a compromise, combining moderate stress (42.7 MPa) and displacement (0.035 mm). This simulation-based study was conducted between January and April 2025 at the Department of Orthodontics, Kocaeli Health and Technology University. A standardized 3D mandibular bone model (2 mm cortical and 13 mm cancellous layers) was constructed, and Ti-6Al-4V miniscrews (1.6 mm × 8 mm) were virtually inserted at 30°, 45°, and 90°. A horizontal orthodontic load of 2 N was applied, and von Mises stress and displacement values were calculated in ANSYS Workbench. Stress patterns were visualized using color-coded maps. The 90° insertion angle generated the highest von Mises stress in cortical bone (50.6 MPa), with a total maximum stress of 58.2 MPa, followed by 45° (42.7 MPa) and 30° (36.4 MPa) insertions (p < 0.001). Stress was predominantly concentrated at the cortical entry point, especially in the 90° model. In terms of displacement, the 90° group exhibited the lowest mean displacement (0.023 ± 0.002 mm), followed by 45° (0.035 ± 0.003 mm) and 30° (0.052 ± 0.004 mm), with statistically significant differences among all groups (p < 0.001). The 45° angle showed a balanced biomechanical profile, combining moderate stress and displacement values, as confirmed by post hoc analysis. From a biomimetics perspective, understanding how insertion angle affects bone response provides insights for designing bio-inspired anchorage systems. By simulating natural stress dissipation, this study demonstrates that insertion angle strongly modulates miniscrew performance. Vertical placement (90°) ensures rigidity but concentrates cortical stress, whereas oblique placement, particularly at 45°, offers a balanced compromise with adequate stability and reduced stress. These results emphasize that beyond material properties, surgical parameters such as insertion angle are critical for clinical success. Full article

With the advancement of intelligent technology and the rise in labor costs, manual identification and cutting of 3D cotton cup indentations can no longer meet modern demands. The increasing variety and shape of 3D cotton cups due to personalized requirements make the use of fixed molds for cutting inefficient, leading to a large number of molds and high costs. Therefore, this paper proposes a UNet-based indentation segmentation algorithm to automatically extract 3D cotton cup indentation data. By incorporating the VGG16 network and Leaky-ReLU activation function into the UNet model, the method improves the model’s generalization capability, convergence speed, detection speed, and reduces the risk of overfitting. Additionally, attention mechanisms and an Atrous Spatial Pyramid Pooling (ASPP) module are introduced to enhance feature extraction, improving the network’s spatial feature extraction ability. Experiments conducted on a self-made 3D cotton cup dataset demonstrate a precision of 99.53%, a recall of 99.69%, a mIoU of 99.18%, and an mPA of 99.73%, meeting practical application requirements. The extracted 3D cotton cup indentation contour data is automatically input into an intelligent CNC cutting machine to cut 3D cotton cup. The cutting results of 400 data points show an 0.20 mm ± 0.42 mm error, meeting the cutting accuracy requirements for flexible material 3D cotton cups. This study may serve as a reference for machine vision, image segmentation, improvements to deep learning architectures, and automated cutting machinery for flexible materials such as fabrics. Full article

The size and mass of fish are crucial parameters in aquaculture management. However, existing research primarily focuses on conducting fish size and mass estimation under ideal conditions, which limits its application in actual aquaculture scenarios with complex water quality and fluctuating lighting. A non-contact size and mass measurement framework is proposed for complex underwater environments, which integrates the improved FishKP-YOLOv11 module based on YOLOv11, stereo vision technology, and a Random Forest model. This framework fuses the detected 2D key points with binocular stereo technology to reconstruct the 3D key point coordinates. Fish size is computed based on these 3D key points, and a Random Forest model establishes a mapping relationship between size and mass. For validating the performance of the framework, a self-constructed grass carp dataset for key point detection is established. The experimental results indicate that the mean average precision (mAP) of FishKP-YOLOv11 surpasses that of diverse versions of YOLOv5–YOLOv12. The mean absolute errors (MAEs) for length and width estimations are 0.35 cm and 0.10 cm, respectively. The MAE for mass estimations is 2.7 g. Therefore, the proposed framework is well suited for application in actual breeding environments. Full article

This research explores a phase-shift-based discontinuous PWM method used for 24 V battery-powered onboard micro inverters, which are critical for thermally limited applications like micromachines, where efficient heat dissipation and compact size are paramount. Discontinuous pulse width modulation (DPWM) reduces switching losses by clamping the phase voltage to the DC bus in order to improve inverter efficiency. Due to the change in power factor at different operating points from motors or the inductor load, the use of only one DPWM method cannot achieve the optimal efficiency of a three-phase voltage source inverter (3ph-VSI). This paper proposes a generalized DPWM method with a continuously adjustable phase shift angle, which extends the six traditional DPWM methods to any type. According to different power factors, the proposed DPWM method is divided into five power factor angle intervals, namely [−90°, −60°], [−60°, −30°], [−30°, 30°], [30°, 60°], and [60°, 90°], and automatically adjusts the phase shift angle to the optimal-efficiency DPWM mode. The power factor is calculated by means of the Synchronous Reference Frame Phase-Locked Loop (SRF-PLL) method. The switching losses and harmonic characteristics of the proposed DPWM are analyzed, and finally, a 24 V onboard 3ph-VSI experimental platform is built. The experimental results show that the efficiency of DPWM methods can be improved by 3–6% and the switching loss can be reduced by 40–50% under different power factors. At the same time, the dynamic performance of the proposed algorithm with a transition state is verified. This method is particularly suitable for miniaturized inverters where efficiency and thermal management are critical. Full article

This study examines the effects of partial wakes caused by upstream turbine yaw control on the trailing-edge noise of a downstream turbine under stable and neutral atmospheric conditions. Using a combined model coupling the unsteady vortex lattice method (UVLM) with the Curl wake model, calibrated with large eddy simulation data, wake behavior and noise characteristics were analyzed for yaw angles from −30° to +30°. Results show that partial wakes slightly raise overall noise levels and lateral asymmetry of trailing-edge noise, while amplitude modulation (AM) strength is more strongly influenced by yaw control. AM varies linearly with wake deflection at moderate yaw angles but behaves nonlinearly beyond a threshold due to large wake deflection and deformation. Findings reveal that yaw control can significantly increase the lateral asymmetry in the AM strength directivity pattern of the downstream turbine, and that AM characteristics depend on the complex interplay between inflow distribution and convective amplification effects, highlighting the importance of accurate wake prediction, along with appropriate consideration of observer point location and blade rotation, for evaluating AM characteristics of a wind turbine influenced by a partial wake. Full article

With the increasing demands of the grid on power electronic converters, active neutral-point-clamped (ANPC) converters have been widely adopted due to their flexible modulation strategies and wide-range power regulation capabilities. To address grid-integration testing requirements for ANPC converters, this paper comparatively studies three electromagnetic transient (EMT) modeling approaches: switch-state prediction method (SPM), associated discrete circuit (ADC), and time-averaged method (TAM). Steady-state and transient simulations reveal that the SPM model achieves the highest accuracy (error ≤ 0.018%), while the TAM-based switching function model optimizes the efficiency–accuracy trade-off with 6.4× speedup versus traditional methods and acceptable error (≤2.62%). Consequently, the TAM model is implemented in a real-time hardware-in-the-loop (HIL) platform. Validation under symmetrical/asymmetrical grid faults confirms both the model’s efficacy and the controller’s robust fault ride-through capability. Full article

Pine wilt disease (PWD), transmitted by Monochamus alternatus (JPS), poses a severe threat to global pine forests. Although the entomopathogenic fungi Beauveria bassiana (Bb) and Metarhizium anisopliae (Ma) represent environmentally friendly biocontrol alternatives, their practical application is limited by inconsistent field performance and an incomplete understanding of host–pathogen interactions. We employed dual RNA-seq at the critical 48 h infection time point to systematically compare the transcriptional responses between JPS and Bb/Ma during infection. Key findings revealed distinct infection strategies: Bb preferentially induced autophagy pathways and modulated host carbohydrate metabolism to facilitate nutrient acquisition, triggering corresponding tissue degradation responses in JPS. In contrast, Ma primarily co-opted host amino acid and sugar metabolic pathways for biosynthetic processes, eliciting a stronger immune defense activation in JPS. Notably, the mTOR signaling pathway was identified as a key regulator of the differential host responses to various entomopathogenic fungi. Further functional validation-specifically, the application of a chemical inhibitor and RNAi targeting mTOR in JPS-confirmed that mTOR inhibition selectively enhanced Bb-induced mortality in JPS without affecting Ma virulence. Our findings reveal the molecular determinants of host–pathogen specificity in PWD biological control and indicate that mTOR regulation could serve as an effective strategy to improve fungal pesticide performance. Full article

Semantic segmentation of LiDAR point clouds in powerline corridor environments is crucial for the intelligent inspection and maintenance of power infrastructure. However, existing deep learning methods often underperform in such scenarios due to severe class imbalance, sparse and long-range structures, and complex elevation variations. We propose EMPower-Net, an Elevation-Aware Multi-Resolution Network, which integrates an Elevation Distribution (ED) module to enhance vertical geometric awareness and a Multi-Resolution (MR) module to enhance segmentation accuracy for corridor structures with varying object scales. Experiments on real-world datasets from Yunnan and Guangdong show that EMPower-Net outperforms state-of-the-art baselines, especially in recognizing power lines and towers with high structural fidelity under occlusion and dense vegetation. Ablation studies confirm the complementary effects of the MR and ED modules, while transfer learning results reveal strong generalization with minimal performance degradation across different powerline regions. Additional tests on urban datasets indicate that the proposed elevation features are also effective for vertical structure recognition beyond powerline scenarios. Full article

The Automatic Checkout (ACO) task aims to accurately generate complete shopping lists from checkout images. Severe product occlusions, numerous categories, and cluttered layouts impose high demands on detection models’ robustness and generalization. To address these challenges, we propose the Edge-Embedded Multi-Feature Fusion Network (E2MF2Net), which jointly optimizes synthetic image generation and feature modeling. We introduce the Hierarchical Mask-Guided Composition (HMGC) strategy to select natural product poses based on mask compactness, incorporating geometric priors and occlusion tolerance to produce photorealistic, structurally coherent synthetic images. Mask-structure supervision further enhances boundary and spatial awareness. Architecturally, the Edge-Embedded Enhancement Module (E3) embeds salient structural cues to explicitly capture boundary details and facilitate cross-layer edge propagation, while the Multi-Feature Fusion Module (MFF) integrates multi-scale semantic cues, improving feature discriminability. Experiments on the RPC dataset demonstrate that E2MF2Net outperforms state-of-the-art methods, achieving checkout accuracy (cAcc) of 98.52%, 97.95%, 96.52%, and 97.62% on Easy, Medium, Hard, and Average mode, respectively. Notably, it improves by 3.63 percentage points in the heavily occluded Hard mode and exhibits strong robustness and adaptability in incremental learning and domain generalization scenarios. Full article

Transketolase-like protein 1 (TKTL-1) has been implicated in altered cancer metabolism, yet its structure and molecular function remain poorly understood. In this study, we established a homology model of TKTL-1 using multiple templates and validated it through sequence alignment and structural comparison with the canonical transketolase (TKT). Binding-site identification was performed using CASTp, receptor cavity mapping, and blind docking, all of which consistently pointed to a conserved region involving interactive residues shared between TKT and TKTL-1. Comparative docking revealed the reduced affinity of TKTL-1 for TDP, supporting functional divergence between TKTL-1 and TKT. We further analyzed conserved residues and receptor surfaces, which enabled us to propose predictive scaffolds as potential modulators of TKTL-1. While these scaffolds remain theoretical, they provide a computational framework to guide future pharmacophore modeling, molecular dynamics simulations, and experimental validation. Together, our study highlights the structural features of TKTL-1, establishes its key differences from TKT, and lays the groundwork for future drug discovery efforts targeting cancer metabolism. Full article