Visual Model Research Articles

Investigating the Effect and Potential Mechanism of Rhamnetin 3-O-α-Rhamnoside on Acute Liver Injury In Vivo and In Vitro

Background/Objectives: Rhamnetin 3-O-α-rhamnoside (ARR) is a major flavonoid of the herb Loranthus tanakae Franch. & Sav., which has been used for treating liver diseases in China. However, the protective effect of ARR on the liver has not been reported. Methods: Zebrafish larvae were used as a visual animal model, and liver injury was induced by thioacetamide (TAA) for an acute liver injury (ALI) model. The hepatoprotective activity of ARR was evaluated by assessing liver morphology, liver function indices, oxidative stress, and the mRNA expression levels of inflammation-related genes in the zebrafish model. Additionally, the ROS level, inflammatory factors, and protein expression related to the IKKβ/NF-κB signaling pathway were measured to investigate a potential mechanism of ARR in HepG2 cells. Results: ARR ameliorated TAA-induced growth retardation, reduced liver injury phenotypes, and decreased oxidative stress in the zebrafish. ARR was also able to lower ROS levels in HepG2 cells, effectively inhibit the overactivation of the IKKβ/NF-κB signaling pathway in pathological conditions, inhibit NF-κB p65 translocation from the cytoplasm to the nucleus, and reduce the release of intracellular inflammatory factors. Conclusions: ARR showed significant protective activity against TAA-induced liver injury in in vivo and in vitro models, and its potential mechanism was closely related to the IKKβ/NF-κB signaling pathway.

Visual collective behaviors on spherical robots.

The implementation of collective motion, traditionally, disregard the limited sensing capabilities of an individual, to instead assuming an omniscient perception of the environment. This study implements a visual flocking model in a ``robot-in-the-loop'' approach to reproduce these behaviors with a flock composed of 10 independent spherical robots. The model achieves robotic collective motion by only using panoramic visual information of each robot, such as retinal position, optical size and optic flow of the neighboring robots. We introduce a virtual anchor to confine the collective robotic movements so to avoid wall interactions. For the first time, a simple visual robot-in-the-loop approach succeed in reproducing several collective motion phases, in particular, swarming, and milling. Another milestone achieved with by this model is bridging the gap between simulation and physical experiments by demonstrating nearly identical behaviors in both environments with the same visual model. To conclude, we show that our minimal visual collective motion model is sufficient to recreate most collective behaviors on a robot-in-the-loop system that is scalable, behaves as numerical simulations predict and is easily comparable to traditional models.

Clinical characteristics and influencing factors of hyperkyphosis in a Chinese Cohort with axial spondyloarthritis: a multicentre retrospective study.

This study aimed to investigate the clinical characteristics and influencing factors of axial spondyloarthritis (axSpA) hyperkyphosis in a Chinese cohort. A cross-sectional study was conducted on 607 patients with axSpA attending 12 hospitals across 11 centers from March 2022 to March 2024.Univariate and multivariate logistic regression analyses were used to explore the relevant influencing factors of hyperkyphosis. A nomogram model for impact factor visualisation and spearman correlation analysis was used to analyze the relationship between the influencing factors. Multivariable logistic regression revealed that male sex,disease duration,patient global assessment (PGA),erythrocyte sedimentation rate (ESR),the modified stoke spondylitis score（mSASSS）,pharmacological treatment, and ankylosing spondylitis disease activity score-c-reactive protein (ASDAS-CRP) significantly influenced hyperkyphosis(all P < 0.05).Based on the results of the multivariate regression analysis, we constructed a nomogram model for clinical evaluation with an AUC of 0.98 and an accuracy of 0.95.Spearman correlation analysis showed a positive correlation between the spondyloarthritis international society health index (ASAS-HI) and mSASSS (R=0.16, P<0.001), while pharmacological treatment was negatively correlated with disease activity and mSASSS (R=-0.24, -0.18, P<0.001). Controlling disease activity in the clinic is crucial. Active pharmacological treatment should be employed to delay radiological progression, enhance patients' ASAS-HI, psychological status, and physical functioning. Additionally, strict smoking cessation and weight control are recommended to reduce disability.

A Bio-Inspired Visual Neural Model for Robustly and Steadily Detecting Motion Directions of Translating Objects Against Variable Contrast in the Figure-Ground and Noise Interference

(1) Background: At present, the bio-inspired visual neural models have made significant achievements in detecting the motion direction of the translating object. Variable contrast in the figure-ground and environmental noise interference, however, have a strong influence on the existing model. The responses of the lobula plate tangential cell (LPTC) neurons of Drosophila are robust and stable in the face of variable contrast in the figure-ground and environmental noise interference, which provides an excellent paradigm for addressing these challenges. (2) Methods: To resolve these challenges, we propose a bio-inspired visual neural model, which consists of four stages. Firstly, the photoreceptors (R1–R6) are utilized to perceive the change in luminance. Secondly, the change in luminance is divided into parallel ON and OFF pathways based on the lamina monopolar cell (LMC), and the spatial denoising and the spatio-temporal lateral inhibition (LI) mechanisms can suppress environmental noise and improve motion boundaries, respectively. Thirdly, the non-linear instantaneous feedback mechanism in divisive contrast normalization is adopted to reduce local contrast sensitivity; further, the parallel ON and OFF contrast pathways are activated. Finally, the parallel motion and contrast pathways converge on the LPTC in the lobula complex. (3) Results: By comparing numerous experimental simulations with state-of-the-art (SotA) bio-inspired models, we can draw four conclusions. Firstly, the effectiveness of the contrast neural computation and the spatial denoising mechanism is verified by the ablation study. Secondly, this model can robustly detect the motion direction of the translating object against variable contrast in the figure-ground and environmental noise interference. Specifically, the average detection success rate of the proposed bio-inspired model under the pure and real-world complex noise datasets was increased by 5.38% and 5.30%. Thirdly, this model can effectively reduce the fluctuation in this model response against variable contrast in the figure-ground and environmental noise interference, which shows the stability of this model; specifically, the average inter-quartile range of the coefficient of variation in the proposed bio-inspired model under the pure and real-world complex noise datasets was reduced by 38.77% and 47.84%, respectively. The average decline ratio of the sum of the coefficient of variation in the proposed bio-inspired model under the pure and real-world complex noise datasets was 57.03% and 67.47%, respectively. Finally, the robustness and stability of this model are further verified by comparing other early visual pre-processing mechanisms and engineering denoising methods. (4) Conclusions: This model can robustly and steadily detect the motion direction of the translating object under variable contrast in the figure-ground and environmental noise interference.

Establishment and validation of a nomogram for coronary artery lesions in children with Kawasaki disease

BackgroundThe nomogram is a powerful and robust tool in disease risk prediction that summarizes complex variables into a visual model that is interpretable with a quantified risk probability. In the current study, a nomogram was developed to predict the occurrence of coronary artery lesions (CALs) among patients with Kawasaki disease (KD). This is especially valuable in the early identification of the risk of CALs, which will lead to proper diagnosis and treatment to reduce their associated complications.MethodsRetrospective clinical data of 677 children diagnosed with KD who were treated in the Children's Hospital Affiliated with Shandong University were analyzed. All the participants were divided into the CAL group and no CAL group according to their coronary echocardiography results. Least absolute shrinkage and selection operator (LASSO) regression was applied for the identification of the most informative predictors of CAL. Based on this, a nomogram was developed for accurate risk estimation.ResultsThe data were divided into a training set and a validation set. Receiver operating characteristic analysis, calibration curves, and decision curve analysis all supported the high accuracy and clinical utility of this model. LASSO regression highlighted five key predictors: sodium, hemoglobin, platelet count, D-dimer, and cystatin C. A nomogram based on these predictors was established and successfully validated in both datasets. In the training set, the AUC was 0.819 and in the validation set it was 0.844. The C-index of the calibration curve in the training set was 0.820, while in the validation set it was 0.844. In the decision curve analysis, the predictive benefit of the model was greater than zero when the threshold probability was below 95% in the training set and below 92% in the validation set.ConclusionThe predictive factors identified through the LASSO regression approach and the development of the nomogram are important contributions in this respect. This model had a high predictive accuracy and reliability for identifying high-risk children in the very early stage of disease with remarkable precision, laying the foundation for personalized treatment strategies and targeted treatment and providing a strong scientific basis for precise therapeutic intervention.

A Novel Approach to Optimize Key Limitations of Azure Kinect DK for Efficient and Precise Leaf Area Measurement

Maize leaf area offers valuable insights into physiological processes, playing a critical role in breeding and guiding agricultural practices. The Azure Kinect DK possesses the real-time capability to capture and analyze the spatial structural features of crops. However, its further application in maize leaf area measurement is constrained by RGB–depth misalignment and limited sensitivity to detailed organ-level features. This study proposed a novel approach to address and optimize the limitations of the Azure Kinect DK through the multimodal coupling of RGB-D data for enhanced organ-level crop phenotyping. To correct RGB–depth misalignment, a unified recalibration method was developed to ensure accurate alignment between RGB and depth data. Furthermore, a semantic information-guided depth inpainting method was proposed, designed to repair void and flying pixels commonly observed in Azure Kinect DK outputs. The semantic information was extracted using a joint YOLOv11-SAM2 model, which utilizes supervised object recognition prompts and advanced visual large models to achieve precise RGB image semantic parsing with minimal manual input. An efficient pixel filter-based depth inpainting algorithm was then designed to inpaint void and flying pixels and restore consistent, high-confidence depth values within semantic regions. A validation of this approach through leaf area measurements in practical maize field applications—challenged by a limited workspace, constrained viewpoints, and environmental variability—demonstrated near-laboratory precision, achieving an MAPE of 6.549%, RMSE of 4.114 cm², MAE of 2.980 cm², and R² of 0.976 across 60 maize leaf samples. By focusing processing efforts on the image level rather than directly on 3D point clouds, this approach markedly enhanced both efficiency and accuracy with the sufficient utilization of the Azure Kinect DK, making it a promising solution for high-throughput 3D crop phenotyping.

Soil stressors on ecophysiology of bauxite mine impacted soil: Heavy metal-acidity-organic matter nexus.

Soil ecophysiology is adversely affected by various environmental hazards, particularly in mining regions. While there has been substantial research on the effects of coal, mica, copper (Cu), and manganese (Mn) mining on soil quality, the impact of bauxite mining operations on nearby soils has largely been overlooked in the literature. Therefore, this study aims to investigate how microbial activity and dynamics are influenced by soil stressors, such as acidity and heavy metals, in areas adjacent to active bauxite mines. Soil samples were collected from three adjacent locations of an active bauxite mine area at distances of<100m (S1), 100-500m (S2), and>500m (S3). The samples contained chromium (Cr), copper (Cu), nickel (Ni), lead (Pb), zinc (Zn), manganese (Mn), and cadmium (Cd), as well as elevated acidity and aluminum (Al). These conditions adversely affected the soil microbial indicators, including fluorescein diacetate (FDA), microbial biomass carbon (MBC), and enzyme activity. The highest concentrations of labile metals (i.e., water-soluble and exchangeable) were found in soil mixed with mining waste (S1), whereas acidity and Al were highest in sparsely vegetated soil (S3). Total acidity, total potential acidity, pH-dependent acidity, and Al were significantly positively correlated. Moreover, the significant positive correlation among organic carbon (OC), acidity, Al, and microbial properties (FDA, MBC, and microbial enzymes) suggests a potential effect of OC in mitigating acidity in S3. The ratios of microbial properties with OC depicted a significant negative correlation with acidity and Al fraction, denoting that acidity and Al posed a deleterious effect on soil microbial health. The similarity percentage analysis identified acid phosphatase as the key enzyme accounting for ∼78% of the observed differences in enzyme composition across the sites. Visual MINTEQ modeling revealed that the sites were saturated with different Al-bearing minerals. Pollution load index (PI) and the geo-accumulation index (Igeo) values identified the region as heavily contaminated (PI>1). Finally, the health risk analysis revealed that Ni posed a potential carcinogenic risk for both adults and children.

AGSAM-Net: UAV route planning and visual guidance model for bridge surface defect detection

AGSAM-Net: UAV route planning and visual guidance model for bridge surface defect detection

VFMDet: A Visual Filtering Mechanism-Based SAR Ship Detection Model for Complex Environment

VFMDet: A Visual Filtering Mechanism-Based SAR Ship Detection Model for Complex Environment

VmambaIR: Visual State Space Model for Image Restoration

VmambaIR: Visual State Space Model for Image Restoration

Fusion of Visual Attention Model and SVM for Sentiment Analysis of Planar Images

Fusion of Visual Attention Model and SVM for Sentiment Analysis of Planar Images

Sub-lethal pesticide exposure interferes with honey bee memory of learnt colours.

Neonicotinoid pesticide use has increased around the world despite accumulating evidence of their potential detrimental sub-lethal effects on the behaviour and physiology of bees, and its contribution to the global decline in bee health. Whilst flower colour is considered as one of the most important signals for foraging honey bees (Apis mellifera), the effects of pesticides on colour vision and memory retention in a natural setting remain unknown. We trained free flying honey bee foragers by presenting artificial yellow flower feeder, to an unscented artificial flower patch with 6 different flower colours to investigate if sub-lethal levels of imidacloprid would disrupt the acquired association made between the yellow flower colour from the feeder and food reward. We found that for doses higher than 4% of LD50 value, the foraging honey bees no longer preferentially visited the yellow flowers within the flower patch and instead, we suspect, reverted back to baseline foraging preferences, with a complete loss of the yellow preference. Our honey bee colour vision modelling indicates that discriminating the yellow colour from the rest should have been easy cognitive task. Pesticide exposure also resulted in a significant increase in Lop1, UVop, and Blop, and a decrease in CaMKII and CREB gene expression. Our results suggest that memory loss is the most plausible mechanism to explain the alteration of bee foraging colour preference. Across bees, colour vision is highly conserved and is essential for efficient pollination services. Therefore, our findings have important implications for ecosystem health and agricultural services world-wide.

Systematic literature review on the effects of blended learning in nursing education.

The study aimed to comprehensively analyze and evaluate the effects of blended learning in nursing education. Blended learning addresses significant challenges in nursing education by offering flexible and adaptive learning environments that support the development of both theoretical knowledge and practical skills. Its importance grew post-COVID-19. Although some reviews have synthesized its effects in nursing education, they often focused on a limited scope and studies, leaving gaps in understanding its broader impact. Systematic literature review. We employed Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines to search for literature published in PubMed, CINAHL, Web of Science, EMBASE, ERIC, Scopus and Cochrane Library from 2000 to June 26, 2024. Two researchers independently screened each article for eligibility and extracted the data. This review included 54 studies and found that blended learning positively influenced nursing students in five key areas: academic performance and skill development, learning engagement and motivation, self-management in learning, psychological and emotional well-being, learning experience and satisfaction. These were captured in a visual model showing the comprehensive impact of blended learning. Critical research gaps were identified, including a lack of randomized controlled trials and limited studies on mental health and clinical thinking abilities. The results of the study showed that the effects of blended learning in nursing education were positive and multi-dimensional. Despite some inconsistencies in findings, blended learning was shown to offer significant advantages over traditional learning, though further research is required to address areas of uncertainty.

Beware of Validation by Eye: Visual Validation of Linear Trends in Scatterplots.

Visual validation of regression models in scatterplots is a common practice for assessing model quality, yet its efficacy remains unquantified. We conducted two empirical experiments to investigate individuals' ability to visually validate linear regression models (linear trends) and to examine the impact of common visualization designs on validation quality. The first experiment showed that the level of accuracy for visual estimation of slope (i.e., fitting a line to data) is higher than for visual validation of slope (i.e., accepting a shown line). Notably, we found bias toward slopes that are "too steep" in both cases. This lead to novel insights that participants naturally assessed regression with orthogonal distances between the points and the line (i.e., ODR regression) rather than the common vertical distances (OLS regression). In the second experiment, we investigated whether incorporating common designs for regression visualization (error lines, bounding boxes, and confidence intervals) would improve visual validation. Even though error lines reduced validation bias, results failed to show the desired improvements in accuracy for any design. Overall, our findings suggest caution in using visual model validation for linear trends in scatterplots.

FusionMamba: dynamic feature enhancement for multimodal image fusion with Mamba

Multimodal image fusion aims to integrate information from different imaging techniques to produce a comprehensive, detail-rich single image for downstream vision tasks. Existing methods based on local convolutional neural networks (CNNs) struggle to capture global features efficiently, while Transformer-based models are computationally expensive, although they excel at global modeling. Mamba addresses these limitations by leveraging selective structured state space models (S4) to effectively handle long-range dependencies while maintaining linear complexity. In this paper, we propose FusionMamba, a novel dynamic feature enhancement framework that aims to overcome the challenges faced by CNNs and Vision Transformers (ViTs) in computer vision tasks. The framework improves the visual state-space model Mamba by integrating dynamic convolution and channel attention mechanisms, which not only retains its powerful global feature modeling capability, but also greatly reduces redundancy and enhances the expressiveness of local features. In addition, we have developed a new module called the dynamic feature fusion module (DFFM). It combines the dynamic feature enhancement module (DFEM) for texture enhancement and disparity perception with the cross-modal fusion Mamba module (CMFM), which focuses on enhancing the inter-modal correlation while suppressing redundant information. Experiments show that FusionMamba achieves state-of-the-art performance in a variety of multimodal image fusion tasks as well as downstream experiments, demonstrating its broad applicability and superiority.

Perceptual image hashing by exploiting a visual attention model along with discrete wavelet transform and discrete cosine transform

Perceptual image hashing by exploiting a visual attention model along with discrete wavelet transform and discrete cosine transform

Models for predicting vaginal birth after cesarean section: scoping review

BackgroundWomen who are pregnant again after a prior cesarean section are faced with the choice between a vaginal trial and another cesarean section. Vaginal delivery is safer for mothers and babies, but face the risk of trial labor failure. Predictive models can evaluate the success rate of vaginal trial labor after cesarean section, which will help obstetricians and pregnant women choose the appropriate delivery method.ObjectiveTo review the existing prediction models of vaginal delivery after cesaean.MethodsSeven databases, including CNKI, Wanfang Data, Chinese Science and Technology Periodical Database, China Biomedical Literature Database, PubMed, Embase, and Web of Science, were searched for studies on the predictive model of VBAC from inception to July 20, 2022. Two researchers independently screened the literature and extracted the data. The risk of bias and applicability of the included researches was evaluated using the Prediction model Risk of Bias Assessment Tool.ResultsTwenty-six studies that covered 26 models were included. The overall property of the included models was good, but validation of the included models was insufficient. The methodological quality of the included studies was generally low, with 3 studies rated as having a low risk of bias and 23 studies rated as having a high risk of bias. The main predictors in the models were the Bishop score, history of vaginal delivery, neonatal weight, maternal age, and BMI.ConclusionsAlthough a variety of prediction models have been developed globally, the methodology of these studies has limitations and the models have not been adequately validated. In the future, more prospective and high-quality research is needed to develop visual models to serve clinical work more effectively and conveniently. Obstetricians or midwives could use predictive models to help a woman choose the right delivery method.

Single visual model based on transformer for digital instrument reading recognition

Abstract Digital instrument reading recognition (DIRR) technology is crucial for industrial digital transformation and the advancement of industrialisation. However, digital instruments differ in character fonts, styles, spacing, and aspect ratios, as well as the scarcity of data pose significant challenges to current recognition technologies. To address these challenges, this study proposed a novel single visual model based on transformer for digital instrument recognition (SVDIR). The SVDIR model primarily comprised a scaled cosine attention mechanism (SC-attention) and a local Transformer block. First, the SC-attention was designed to calculate the cosine similarity of two image patches. It rendered the attention calculation independent of the input amplitude and produced milder attention weights to alleviate overconcentration issues. Second, a local Transformer block module was proposed for extracting the internal stroke features and dependencies between character components. Fine-grained characteristic features were obtained using this method. In addition, a post-norm structure was introduced into the local Transformer block module to reduce the accumulation of activation values following the deepening of the network. Finally, experimental results demonstrated the effectiveness and superiority of the proposed model on two digital instrument datasets.

Visual Story Generation Model Guided by Multi Granularity Image Information

Visual story generation, which involves generating short stories from sequential images, has become a core task at the intersection of computer vision and natural language processing. However, existing methods suffer from a bias in the concept predicates predicted, leading to a semantic gap between the generated stories and the images. This paper proposes a novel visual story generation model that utilizes muliti granularity image information to guide the generation process and correct the bias in concept predicates, resulting in more image-consistent stories. The proposed model consists of two stages: In the first stage, a set of concepts predicates is predicted from the image and enriched with external knowledge, and the most suitable concepts for story generation are selected. In the second stage, fine-grained image information are utilized to integrate image information into the story generation module, improving the bias in concept predicates. The image theme information and the generated results of previous moments are used as prompts to guide the story generation module. Experimental results show that the proposed model outperforms baseline models in all evaluation metrics. Specifically, the BLEU-1, BLEU-2, BLEU-3, and BLEU-4 metrics are improved by 4.0, 3.8, 3.02, and 1.98 percentage points, respectively, and the METEOR metric is improved by 1.4 percentage points. The generated stories are more consistent with the image content, maintain a consistent theme, and enhance coherence between contexts.

Method for Recognizing Disordered Sugarcane Stacking Based on Improved YOLOv8n

In order to enhance the efficiency and precision of grab-type planting operations for disordered stacked sugarcane, and to achieve rapid deployment of the visual detection model on automatic sugarcane seed-cane planters, this study proposes a sugarcane detection algorithm based on an improved YOLOv8n model. Firstly, the backbone network of YOLOv8n is replaced with VanillaNet to optimize feature extraction capability and computational efficiency; the InceptionNeXt deep convolutional structure is integrated, utilizing its multi-scale processing feature to enhance the model’s ability to recognize sugarcane of different shapes and sizes. Secondly, the ECA attention mechanism is incorporated in the feature fusion module C2F to further enhance the recognition model’s capability to capture key features of sugarcane. The MPDIOU loss function is employed to improve the resolution of recognizing overlapping sugarcane, reducing misidentification and missed detection. Experimental results show that the improved YOLOv8n model achieves 96% and 71.5% in mAP@0.5 and mAP@0.5:0.95 respectively, which are increases of 5.1 and 6.4 percentage points compared to the original YOLOv8n model; moreover, compared to the currently popular Faster-RCNN, SSD, and other YOLO series object detection models, it not only improves detection accuracy but also significantly reduces the number of model parameters. The research results provide technical support for subsequent sugarcane grab-type planting recognition and mobile deployment.