Search Results (1,054)

Collateral status is an important therapeutic indicator for acute ischemic stroke (AIS), yet visual collateral grading remains subjective and suffers from inter-observer variability. To address this limitation, this study automatically extracted binarized vascular morphological features from CTA images and developed a convolutional neural network (CNN) for automated collateral classification. Performance trends were systematically analyzed across diverse hyperparameter combinations to meet different clinical decision needs. A total of 157 AIS patients (median age 65 [⁵⁷,⁵⁸,⁵⁹,⁶⁰,⁶¹,⁶²,⁶³,⁶⁴,⁶⁵,⁶⁶,⁶⁷,⁶⁸,⁶⁹,⁷⁰,⁷¹,⁷²,⁷³,⁷⁴] years; 61.8% were male) were retrospectively enrolled and stratified by Menon score into good (3–5, n = 117) and poor (0–2, n = 40) collateral groups. A total of 192 architectures were established, and three representative model tendencies emerged: a sensitivity-oriented model (AUC = 0.773; sensitivity = 87.18%; specificity = 65.00%), a balanced model (AUC = 0.768; sensitivity = 72.65%; specificity = 77.50%), and a specificity-oriented model (AUC = 0.753; sensitivity = 63.25%; specificity = 85.00%). These results demonstrate that kernel size, the number of filters in the first layer, and the number of convolutional layers are key determinants of performance directionality, allowing tailored model selection depending on clinical requirements. This work highlights the feasibility of CTA-based automated collateral classification and provides a systematic framework for developing models optimized for sensitivity, specificity, or balanced decision-making. The findings may serve as a reference for clinical model deployment and have potential for integration into multi-objective AI systems for endovascular thrombectomy patient triage. Full article

In object detection, boundary blurring caused by occlusion and background interference often hinders effective feature extraction. To address this challenge, we propose Edge Aware-YOLO, a novel framework designed to enhance edge awareness and efficient feature fusion. Our method integrates three key contributions. First, the Variable Sobel Compact Inverted Block (VSCIB) employs convolution kernels with adjustable orientation and size, enabling robust multi-scale edge adaptation. Second, the Spatial Pyramid Shared Convolution (SPSC) replaces standard pooling with shared dilated convolutions, minimizing detail loss during feature reconstruction. Finally, the Efficient Downsampling Convolution (EDC) utilizes a dual-branch architecture to balance channel compression with semantic preservation. Extensive evaluations on public datasets demonstrate that Edge Aware-YOLO significantly outperforms state-of-the-art models. On MS COCO, it achieves 56.3% mAP₅₀ and 40.5% mAP_50–95 (gains of 1.5% and 1.0%) with only 2.4M parameters and 5.8 GFLOPs, surpassing advanced models like YOLOv11. Full article

To achieve accurate voltage sag prediction and early warning, thereby improving power quality, a hybrid voltage sag prediction framework is proposed by integrating Kernel Entropy Component Analysis (KECA) with an improved Northern Goshawk Optimization (NGO) algorithm for hyperparameter tuning of a CNN-LSTM model. First, to address the limitations of the original NGO, such as proneness to falling into local optima and high randomness of the initial population distribution, a refraction-opposition-based learning mechanism is introduced to enhance population diversity and expand the search space. Furthermore, a sine–cosine strategy (SCA) with nonlinear weight coefficients is integrated into the exploration phase to dynamically adjust the search step size, optimizing the balance between global exploration and local exploitation, thereby boosting convergence speed and accuracy. The improved algorithm (SCNGO) is then utilized to optimize the hyperparameters of the CNN-LSTM model. Second, KECA is applied to voltage-sag-related data to extract key features and eliminate redundant information, and the resulting dimensionally reduced data are fed as input to the SCNGO-CNN-LSTM model to further improve prediction performance. Experimental results demonstrate that the SCNGO-CNN-LSTM model outperforms other comparative models significantly across multiple evaluation metrics. Compared with NGO-CNN-LSTM, GWO-CNN-LSTM, and the original CNN-LSTM, the proposed method achieves a mean squared error (MSE) reduction of 53.45%, 44.68%, and 66.76%, respectively. The corresponding root mean squared error (RMSE) is decreased by 25.33%, 18.61%, and 36.92%, while the mean absolute error (MAE) is reduced by 81.23%, 77.04%, and 86.06%, respectively. These results confirm that the proposed framework exhibits superior feature representation capability and significantly improves voltage sag prediction accuracy. Full article

To maintain reliability, safety, and sustainability in power transmission, insulator defect detection has become a critical task in power line inspection. Due to the complex backgrounds and small defect sizes encountered in insulator defect images, issues such as false detections and missed detections often occur. The existing You Only Look Once (YOLO) object detection algorithm is currently the mainstream method for image-based insulator defect detection in power lines. However, existing models suffer from low detection accuracy. To address this issue, this paper presents an improved YOLOv5-based MC-YOLO insulator detection algorithm. To effectively extract multi-scale information and enhance the model’s ability to represent feature information, a multi-scale attention convolutional fusion (MACF) module incorporating an attention mechanism is proposed. This module utilises parallel convolutions with different kernel sizes to effectively extract features at various scales and highlights the feature representation of key targets through the attention mechanism, thereby improving the detection accuracy. Additionally, a cross-context feature fusion module (CCFM) is designed, where shallow features gain partial deep semantic supplementation and deep features absorb shallow spatial information, achieving bidirectional information flow. Furthermore, the Spatial-Channel Dual Attention Module (SCDAM) is introduced into CCFM. By incorporating a dynamic attention-guided bidirectional cross-fusion mechanism, it effectively resolves the feature deviation between shallow details and deep semantics during multi-scale feature fusion. The experimental results show that the MC-YOLO algorithm achieves an mAP@0.5 of 67.4% on the dataset used in this study, which is a 4.1% improvement over the original YOLOv5. Although the FPS is slightly reduced compared to the original model, it remains practical and capable of rapidly and accurately detecting insulator defects. Full article

Coastal wetland vegetation exhibits pronounced spectral mixing, complex mosaic spatial patterns, and small target sizes, posing considerable challenges for fine-grained classification in high-resolution UAV imagery. At present, remote sensing classification of ground objects based on deep learning mainly relies on spectral and structural features, while the frequency domain features of ground objects are not fully considered. To address these issues, this study proposes a vegetation classification model that integrates spatial-domain and frequency-domain features. The model enhances global contextual modeling through a large-kernel convolution branch, while a frequency-domain interaction branch separates and fuses low-frequency structural information with high-frequency details. In addition, a shallow auxiliary supervision module is introduced to improve local detail learning and stabilize training. With a compact parameter scale suitable for real-world deployment, the proposed framework effectively adapts to high-resolution remote sensing scenarios. Experiments on typical coastal wetland vegetation including Reeds, Spartina alterniflora, and Suaeda salsa demonstrate that the proposed method consistently outperforms representative segmentation models such as UNet, DeepLabV3, TransUNet, SegFormer, D-LinkNet, and MCCA across multiple metrics including Accuracy, Recall, F1 Score, and mIoU. Overall, the results show that the proposed model effectively addresses the challenges of subtle spectral differences, pervasive species mixture, and intricate structural details, offering a robust and efficient solution for UAV-based wetland vegetation mapping and ecological monitoring. Full article

In the field management of maize, phenomena such as missed sowing and empty seedlings directly affect the final yield. By implementing seedling replenishment activities and promptly evaluating seedling growth, maize output can be increased by improving seedling survival rates. To address the challenges posed by complex field environments (including varying light conditions, weeds, and foreign objects), as well as the performance limitations of model deployment on resource-constrained devices, this study proposes a Lightweight Real-Time You Only Look Once (LRT-YOLO) model. This model builds upon the You Only Look Once version 11n (YOLOv11n) framework by designing a lightweight, optimized feature architecture (OF) that enables the model to focus on the characteristics of small to medium-sized maize seedlings. The feature fusion network incorporates two key modules: the Feature Complementary Mapping Module (FCM) and the Multi-Kernel Perception Module (MKP). The FCM captures global features of maize seedlings through multi-scale interactive learning, while the MKP enhances the network’s ability to learn multi-scale features by combining different convolution kernels with pointwise convolution. In the detection head component, the introduction of an NMS-free design philosophy has significantly enhanced the model’s detection performance while simultaneously reducing its inference time. The experiments show that the $mA{P}_{50}$ and $mA{P}_{50:95}$ of the LRT-YOLO model reached 95.9% and 63.6%, respectively. The model has only 0.86M parameters and a size of just 2.35 M, representing reductions of 66.67% and 54.89% in the number of parameters and model size compared to YOLOv11n. To enable mobile deployment in field environments, this study integrates the LRT-YOLO model with the ByteTrack multi-object tracking algorithm and deploys it on the NVIDIA Jetson AGX Orin platform, utilizing OpenCV tools to achieve real-time visualization of maize seedling tracking and counting. Experiments demonstrate that the frame rate (FPS) achieved with TensorRT acceleration reached 23.49, while the inference time decreased by 38.93%. Regarding counting performance, when tested using static image data, the coefficient of determination ($R^2$) and root mean square error (RMSE) were 0.988 and 5.874, respectively. The cross-line counting method was applied to test the video data, resulting in an $R^2$ of 0.971 and an RMSE of 16.912, respectively. Experimental results show that the proposed method demonstrates efficient performance on edge devices, providing robust technical support for the rapid, non-destructive counting of maize seedlings in field environments. Full article

Magnetic nanoparticles (MNPs) are widely applied in biomedicine, including bioimaging, drug delivery, and cell tracking. As central mediators of immune surveillance, macrophages phagocytize foreign substances, rendering their interactions with MNPs particularly consequential. During MNP uptake, macrophages undergo cytoplasmic remodeling that can lead to morphological alterations. Although prior studies have predominantly focused on MNP uptake efficiency and cytotoxicity, systematic quantitative assessments of macrophage morphological alterations following MNP treatment remain scarce. In this study, phase-contrast microscopy images of macrophages before and after MNP treatment were analyzed using unsupervised variational autoencoder (VAE)-based frameworks. Specifically, the β-VAE, β-total correlation VAE, and multi-encoder VAE frameworks were employed to extract latent representations of cellular morphology. The analysis revealed that MNP-treated macrophages exhibited pronounced structural alterations, including membrane expansion, central density shifts, and shape distortions. These findings were further substantiated through quantitative evaluations, including effect size analysis, kernel density estimation, latent traversal, and difference mapping. Collectively, these results demonstrate that VAE-based unsupervised learning provides a robust framework for detecting subtle morphological responses of macrophages to nanoparticle exposure and highlights its broader applicability across varied cell types, treatment conditions, and imaging platforms. Full article

In the context of ship target detection with Synthetic Aperture Radar (SAR) images, misdetection and missed detection are often caused by complex background interference and the variability in target size. To address these challenges, this paper proposes an innovative method based on image enhancement and feature fusion to reduce background noise and effectively handle the detection confusion caused by differences in ship sizes. Firstly, a feature-aware enhancement network is introduced, which preserves and strengthens the edge information of the target objects. Secondly, during the feature extraction phase, a dynamic hierarchical extraction module is proposed, significantly improving the feature capture ability of convolutional neural networks and overcoming the limitations of traditional fixed kernel receptive fields. Finally, a feature fusion module based on attention gating is employed to fully leverage the complementary information between the original and enhanced images, achieving precise modeling and efficient fusion of inter-feature correlations. The proposed method is integrated with the YOLOv8 detection framework for target detection. Experimental results in the publicly available SSDD and HRSID datasets demonstrate detection accuracies of 97.9% and 93.2%, respectively, thus validating the superiority and robustness of the proposed method. Full article

An age-distributed immuno-epidemiological model with information-based vaccination proposed in this work represents a system of integro-differential equations with compartments for the numbers of susceptible individuals, infected individuals, vaccinated individuals, and recovered individuals. This model describes the influence of vaccination decisions on epidemic progression in different age groups. In a particular case of the model without age distribution, we determine the basic reproduction number and the final size of epidemic, that is, the limiting number of susceptible individuals at asymptotically large time. Moreover, we study the existence and uniqueness of a positive solution for the age-structured model. Numerical simulations show that the information-based vaccination acceptance can significantly influence the epidemic progression. Though the initial stage of epidemic progression is the same for all memory kernels, as the epidemic progresses and more information about the disease becomes available, further epidemic progression strongly depends on the memory effect. A short-range memory kernel appears to be more effective in restraining the epidemic outbreaks because it allows for more responsive and adaptive vaccination decisions based on the most recent information about the disease. Additionally, the simulation results suggest that relying on either a responsive vaccination approach or a highly effective vaccine alone may be insufficient to significantly reduce the epidemic size and prevent large outbreaks. Both factors are necessary to achieve substantial epidemic control. Moreover, the impacts of the age-dependent initial susceptible population and the age-dependent memory kernel are studied through numerical simulation of the age-dependent model. Full article

The stability of nanolubricants is critical for ensuring effective performance under extreme pressure (EP) conditions, where severe boundary lubrication governs friction and wear behaviour. This study examines palm kernel oil (PKO)-based nanolubricants enhanced with carbon graphene (CG), hexagonal boron nitride (hBN), and molybdenum disulfide (MoS₂), with and without oleic acid (OA) as a surfactant. OA incorporation improved CG dispersion stability, reducing agglomerate size by 30.4% (17.61 μm to 12.23 μm) and increasing the viscosity index from ~176 to 188, compared to 152 for the commercial hydrogen engine oil baseline. Under EP conditions, PKO + CG + OA achieved a 51.7% reduction in the coefficient of friction (0.58 to 0.28) and 18.2% improvement in weld load resistance, while wear scar diameter decreased by 13.4%. Surface and elemental analyses indicated the formation of a composite tribofilm containing oxide species, graphene platelets, and carboxylate-derived compounds from OA, consistent with iron–oleate-like chemistry that enhances load-carrying capacity and wear protection. These findings demonstrate the potential of OA-assisted PKO nanolubricants as sustainable, high-performance formulations for extreme pressure boundary lubrication, contributing to the advancement of green tribology. Full article

Body weight and size are critical indicators of sow health and reproductive performance. Traditional manual measurement methods are not only time-consuming and labor-intensive but also induce stress in sows. To address these limitations, we propose an innovative method for estimating sow body weight and size using 3D point cloud data. Our method began by obtaining point cloud data from depth images captured by an Intel^® RealSense™ D455 camera. First, we used a KPConv segmentation model with a deformable kernel to extract the sow‘s back. The resulting back point cloud then served as the input to a novel dual-branch, multi-output regression model named DbmoNet, which integrates features from both location and feature spaces. We evaluated the method on 2400 samples from three breeds during non-pregnant periods. The KPConv model demonstrated excellent performance, achieving an overall segmentation accuracy (OA) of 99.54%. The proposed DbmoNet model outperformed existing benchmarks, achieving mean absolute percentage errors (MAPEs) of 3.74% for body weight (BW), 3.97% for chest width (CW), 3.33% for hip width (HW), 3.82% for body length (BL), 1.94% for chest height (CH), and 2.43% for hip height (HH). Therefore, this method provides an accurate and efficient tool for non-contact body condition monitoring in intensive sow production. Full article

Aiming at the problems of small-sized ships (such as small patrol boats) in complex open-sea backgrounds, including small sizes, insufficient feature information, and high missed detection rates, this paper proposes a maritime small target image detection algorithm based on the improved YOLOv11n. Firstly, the BIE module is introduced into the neck feature fusion stage of YOLOv11n. Utilizing its dual-branch information interaction design, independent branches for key features of maritime small targets in infrared and visible light images are constructed, enabling the progressive fusion of infrared and visible light target features. Secondly, RepViTBlock is incorporated into the backbone network and combined with the C3k2 module of YOLOv11n to form C3k2-RepViTBlock. Through the lightweight attention mechanism and multi-branch convolution structure, this addresses the insufficient capture of tiny target features by the C3k2 module and enhances the model’s ability to extract local features of maritime small targets. Finally, the ConvAttn module is embedded at the end of the backbone network. With its dynamic small-kernel convolution, it adaptively extracts the contour features of small targets, maintaining the overall model’s light weight while reducing the missed detection rate for maritime small targets. Experiments on a collected infrared and visible light ship image dataset (IVships) and a public dataset (SeaShips) show that, on the basis of increasing only a small number of parameters, the improved algorithm increases the mAP@0.5 by 1.9% and 1.7%, respectively, and the average precision by 2.2% and 2.4%, respectively, compared with the original model, which significantly improves the model’s small target detection capabilities. Full article

With the increasing number of the ocean ships, the demand for synthetic aperture radar (SAR) image ship classification has been much increased. With the development of deep learning, many neural network-based ship classification methods have been presented. However, these networks show unsatisfactory performance on low-quality SAR ship datasets. In this paper, we propose a SAR ship classification method based on dual Metaformer-like networks and learnable decision fusion, which we call LDF-D-MLCNNs. First, we design a Metaformer-like convolutional block to improve learning performance. Secondly, we implement two networks with different kernel sizes and propose the learnable decision fusion module to obtain the final prediction. Kernels of different sizes exhibit diverse extraction capabilities. Experimental results show that the accuracy of the proposed method outperforms many existing SAR ship classification networks. Full article

Essential oils (EOs) are promising bio-preservatives for oilseeds; however, their high volatility and strong aroma limit practical applications. In this study, we developed a dual-size microencapsulated formulation of oregano (Origanum compactum) and myrtle (Myrthus communis) EOs (75:25, w/w) using gelatin–gum arabic complex coacervation, and evaluated its antifungal efficacy and effect on seed viability in peanuts. GC-MS analysis of the EO blend identified carvacrol (33.83%) as the dominant constituent. The microcapsules exhibited an encapsulation efficiency of 83.56% and were produced in a 70% small/30% large particle ratio to ensure both immediate and sustained vapor release. In vapor-phase assays against toxigenic A. flavus (RP-6), both free and encapsulated EOs inhibited fungal growth in a dose-dependent manner and achieved complete suppression at concentrations ≥0.2 µL mL⁻¹, whereas the wall material alone showed no activity. In a 120-day microcosm storage experiment (0.2 mg EO g⁻¹ kernels; 0.96 mg microcapsules g⁻¹), treated peanuts showed an immediate reduction in total fungal load from 3.52 to 1.48 log₁₀ CFU g⁻¹ (≈58%), which stabilized near 1.42–1.43 log₁₀ CFU g⁻¹ up to 90 days, while the control samples increased to 4.25 log₁₀ CFU g⁻¹ by day 120. The formulation effectively suppressed major storage fungi, including Aspergillus sections Flavi and Nigri, Penicillium spp., Rhizopus, Fusarium, and Alternaria. The antioxidant activity (DPPH assay) was retained after encapsulation (IC₅₀: 0.52 mg mL⁻¹ encapsulated vs. 0.58 mg mL⁻¹ free). Germination power remained comparable to the control throughout storage (≈50–52%), indicating no adverse impact on seed viability. These findings demonstrate that vapor-active, dual-size microencapsulation of oregano-myrtle EOs offers a practical and sustainable approach to enhance peanut safety during storage without compromising germination potential. Full article

To promote domestic wheat production in South Korea, four functional colored wheat varieties with varying amylose contents: Ariheuk (AH), Arijinheuk (AJ), Ariheukchal (AC), and Sintong (ST), were developed. This study examined their bread-making performance using whole wheat flour (WWF) milled under different conditions with an ultra-centrifugal mill (sieve openings: 0.5 and 1.0 mm; rotation speeds: 6000 and 14,000 rpm). Four flour samples per variety (FL, FH, CL, CH) were prepared. The median particle size (d50) varied among varieties, with harder kernels (AC, AH) producing larger particles than softer ones (AJ, ST). Smaller sieve openings increased the water and sodium carbonate solvent retention capacity, whereas higher rotation speeds reduced them, indicating less damaged starch. Sodium dodecyl sulfate sedimentation volume was higher in AC and AH, suggesting stronger gluten. Bread made from the group F WWF had higher volume and lower firmness, with AH-FH producing the best bread quality. Total phenolic and anthocyanin content and antioxidant activity were slightly higher in the group F, but markedly lower in the ST. Bread crusts showed increased phenolic and antioxidant activity but decreased anthocyanin content due to heat. Overall, kernel hardness, milling conditions, and amylose content strongly influenced purple WWF quality and bread performance, highlighting the need to optimize milling and formulation strategies. Full article