Search Results (2,946)

Accurate classification of chest X-ray (CXR) images is crucial for diagnosing lung diseases in medical imaging. Existing deep learning models for CXR image classification face challenges in distinguishing non-lung features. In this work, we propose a new segmentation-assisted fusion-based classification method. The method involves two stages: first, we use a lightweight segmentation model, Partial Convolutional Segmentation Network (PCSNet) designed based on an encoder–decoder architecture, to accurately obtain lung masks from CXR images. Then, a fusion of the masked CXR image with the original image enables classification using the improved lightweight ShuffleNetV2 model. The proposed method is trained and evaluated on segmentation datasets including the Montgomery County Dataset (MC) and Shenzhen Hospital Dataset (SH), and classification datasets such as Chest X-Ray Images for Pneumonia (CXIP) and COVIDx. Compared with seven segmentation models (U-Net, Attention-Net, SegNet, FPNNet, DANet, DMNet, and SETR), five classification models (ResNet34, ResNet50, DenseNet121, Swin-Transforms, and ShuffleNetV2), and state-of-the-art methods, our PCSNet model achieved high segmentation performance on CXR images. Compared to the state-of-the-art Attention-Net model, the accuracy of PCSNet increased by 0.19% (98.94% vs. 98.75%), and the boundary accuracy improved by 0.3% (97.86% vs. 97.56%), while requiring 62% fewer parameters. For pneumonia classification using the CXIP dataset, the proposed strategy outperforms the current best model by 0.14% in accuracy (98.55% vs. 98.41%). For COVID-19 classification with the COVIDx dataset, the model reached an accuracy of 97.50%, the absolute improvement in accuracy compared to CovXNet was 0.1%, and clinical metrics demonstrate more significant gains: specificity increased from 94.7% to 99.5%. These results highlight the model’s effectiveness in medical image analysis, demonstrating clinically meaningful improvements over state-of-the-art approaches. Full article

Current mainstream classification methods predominantly employ end-to-end multi-class frameworks. These approaches encounter inherent challenges including high-dimensional feature space complexity, decision boundary ambiguity that escalates with increasing class cardinality, sensitivity to label noise, and limited adaptability to dynamic model expansion. However, human beings may avoid these mistakes naturally. Research indicates that humans subconsciously employ a decision-making process favoring binary outcomes, particularly when responding to questions requiring nuanced differentiation. Intuitively, responding to binary inquiries such as “yes/no” often proves easier for humans than addressing queries of “what/which”. Inspired by the human decision-making hypothesis, we proposes a decision paradigm named the evolutionary binary decision framework (EBDF) centered around binary classification, evolving from traditional multi-classifiers in deep learning. To facilitate this evolution, we leverage the top-N outputs from the traditional multi-class classifier to dynamically steer subsequent binary classifiers, thereby constructing a cascaded decision-making framework that emulates the hierarchical reasoning of a binary decision tree. Theoretically, we demonstrate mathematical proof that by surpassing a certain threshold of the performance of binary classifiers, our framework may outperform traditional multi-classification framework. Furthermore, we conduct experiments utilizing several prominent deep learning models across various image classification datasets. The experimental results indicate significant potential for our strategy to surpass the ceiling in multi-classification performance. Full article

The growing urban population has made densification a key focus of urban development. It is crucial to create an urban planning strategy that understands the environmental, social, and economic effects of densification at both the district and city levels. In Switzerland, densification is a legally binding aim to foster housing and jobs within urban boundaries. The challenge is to accommodate population growth while maintaining a high quality of life. Zurich exemplifies this situation, necessitating the accommodation of approximately 25% of the anticipated increase in both the resident population and associated workplaces, as of 2016. This study examined the effects of urban densification on urban forms and microclimates in the Altstetten–Albisrieden district. It developed five densification scenarios based on current urban initiatives and assessed their impacts. Results showed that the current Building and Zoning Plan provides sufficient capacity to accommodate growth. Strategies such as densifying parcels older than fifty years and adding floors to newer buildings were found to minimally impact existing urban forms. Using the SOLWEIG model in the Urban Multi-scale Environmental Predictor (UMEP), this study simulated mean radiant temperature (Tmrt) in the selected urban areas. The results demonstrated that densification reduced daytime average temperatures by 0.60 °C and diurnal averages by 0.23 °C, but increased average nighttime temperatures by 0.38 °C. This highlights the importance of addressing warm nights. The study concludes that well-planned densification can significantly contribute to urban liveability, emphasising the need for thoughtful building design to improve outdoor thermal comfort. Full article

The accurate extraction of navigational regions from images of navigational waters plays a key role in ensuring on-water safety and the automation of unmanned vessels. Nonetheless, current technological methods encounter significant challenges in addressing fluctuations in water surface illumination, reflective disturbances, and surface undulations, among other disruptions, in turn making it challenging to achieve rapid and precise boundary segmentation. To cope with these challenges, in this paper, we propose a coordinate-aware multi-scale feature network (GASF-ResNet) method for water segmentation. The method integrates the attention module Global Grouping Coordinate Attention (GGCA) in the four downsampling branches of ResNet-50, thus enhancing the model’s ability to capture target features and improving the feature representation. To expand the model’s receptive field and boost its capability in extracting features of multi-scale targets, the Avoidance Spatial Pyramid Pooling (ASPP) technique is used. Combined with multi-scale feature fusion, this effectively enhances the expression of semantic information at different scales and improves the segmentation accuracy of the model in complex water environments. The experimental results show that the average pixel accuracy (mPA) and average intersection and union ratio (mIoU) of the proposed method on the self-made dataset and on the USVInaland unmanned ship dataset are 99.31% and 98.61%, and 98.55% and 99.27%, respectively, significantly better results than those obtained for the existing mainstream models. These results are helpful in overcoming the background interference caused by water surface reflection and uneven lighting in the aquatic environment and in realizing the accurate segmentation of the water area for the safe navigation of unmanned vessels, which is of great value for the stable operation of unmanned vessels in complex environments. Full article

This review investigates recent developments in cardiac mechano-electrical-fluid interaction (MEFI) modeling, with a focus on multiphysics simulation platforms and digital twin frameworks developed between 2015 and 2025. The purpose of the study is to assess how computational modeling methods—particularly finite element and immersed boundary techniques, monolithic and partitioned coupling schemes, and artificial intelligence (AI)-enhanced surrogate modeling—capture the integrated dynamics of cardiac electrophysiology, tissue mechanics, and hemodynamics. The goal is to evaluate the translational potential of MEFI models in clinical applications such as cardiac resynchronization therapy (CRT), arrhythmia classification, atrial fibrillation ablation, and surgical planning. Quantitative results from the literature demonstrate <5% error in pressure–volume loop predictions, >0.90 F1 scores in machine-learning-based arrhythmia detection, and <10% deviation in myocardial strain relative to MRI-based ground truth. These findings highlight both the promise and limitations of current MEFI approaches. While recent advances improve physiological fidelity and predictive accuracy, key challenges remain in achieving multiscale integration, model validation across diverse populations, and real-time clinical applicability. The review concludes by identifying future milestones for clinical translation, including regulatory model certification, standardization of validation protocols, and integration of patient-specific digital twins into electronic health record (EHR) systems. Full article

Air quality forecasts play a crucial role in informing the public about atmospheric pollutant levels that pose risks to human health and the environment. The accuracy of these forecasts strongly depends on the quality and resolution of the input data used in the modelling process. At HungaroMet, the Hungarian Meteorological Service, the CHIMERE chemical transport model is used to provide two-day air quality forecasts for the territory of Hungary. This study compares two configurations of the CHIMERE model: the current operational setup, which uses climatological averages from the LMDz-INCA database for boundary conditions, and a test configuration that incorporates real-time boundary conditions from the CAMS global forecast. The primary objective of this work was to assess how the use of real-time versus climatological boundary conditions affects modelled concentrations of key pollutants, including NO₂, O₃, PM₁₀, and PM_2.5. The model results were evaluated against observational data from the Hungarian Air Quality Monitoring Network using a range of statistical metrics. The results indicate that the use of real-time boundary conditions, particularly for aerosol-type pollutants, improves the accuracy of PM₁₀ forecasts. This improvement is most significant under meteorological conditions that favour the long-range transport of particulate matter, such as during Saharan dust or wildfire episodes. These findings highlight the importance of incorporating dynamic, up-to-date boundary data, especially for particulate matter forecasting—given the increasing frequency of transboundary dust events. Full article

Background/Objective: Precise aneurysm volume measurement offers a transformative edge for risk assessment and treatment planning in clinical settings. Currently, clinical assessments rely heavily on manual review of medical imaging, a process that is time-consuming and prone to inter-observer variability. The widely accepted standard of care primarily focuses on measuring aneurysm diameter at its widest point, providing a limited perspective on aneurysm morphology and lacking efficient methods to measure aneurysm volumes. Yet, volume measurement can offer deeper insight into aneurysm progression and severity. In this study, we propose an automated approach that leverages the strengths of pre-trained neural networks and expert systems to delineate aneurysm boundaries and compute volumes on an unannotated dataset from 60 patients. The dataset includes slice-level start/end annotations for aneurysm but no pixel-wise aorta segmentations. Method: Our method utilizes a pre-trained UNet to automatically locate the aorta, employs SAM2 to track the aorta through vascular irregularities such as aneurysms down to the iliac bifurcation, and finally uses a Long Short-Term Memory (LSTM) network or expert system to identify the beginning and end points of the aneurysm within the aorta. Results: Despite no manual aorta segmentation, our approach achieves promising accuracy, predicting the aneurysm start point with an R² score of 71%, the end point with an R² score of 76%, and the volume with an R² score of 92%. Conclusions: This technique has the potential to facilitate large-scale aneurysm analysis and improve clinical decision-making by reducing dependence on annotated datasets. Full article

Drosophilidae is a relatively small family within Diptera. However, species of this family occupy a wide range of ecological niches and are frequently found in synanthropic habitats. Additionally, some species are known agricultural pests. The dataset is based on collections of Drosophilidae from eleven regions of Russia. The dataset was uploaded to the GBIF platform in 2024. Published sources specifying exact localities and collection dates were also used. The database includes records dating back to 1867, with the majority of specimens collected by the authors between 2001 and 2024. Collection methods included net sweeping and bait trapping. The dataset contains 2830 occurrence records, with a total of 51,006 specimens of Drosophilidae studied. It includes data on 108 species from two subfamilies, covering 49 regions of Russia. Considering additional published sources, 188 species of Drosophilidae are currently known from Russia, with a complete species list provided. Among the most abundant species in the dataset, 10 species are represented by more than 1000 specimens: Drosophila obscura, Scaptodrosophila rufifrons, Drosophila melanogaster, Drosophila phalerata, Drosophila transversa, Drosophila kuntzei, Drosophila histrio, Drosophila testacea, Phortica semivirgo, and Drosophila immigrans. Conversely, 39 species are represented by fewer than 10 specimens in the dataset. Regarding ecological groupings, the most dominant groups are xylosaprobionts (39 species, 40.6%) and mycophages (30 species, 31.3%). Notably, in 2017, the quarantine pest Drosophila suzukii was detected in the European part of Russia. The current knowledge of the Drosophilidae fauna in Russia remains insufficient. Of the 15 regions, only one or two localities are represented in the dataset. The distribution limits and range boundaries of many species remain unknown. Moreover, the local faunas of more than half of Russia’s regions remain unexplored. Full article

In reality, some influencers who publicly share failures can gain more attention, which defies common sense. Existing research primarily focuses on the depth of self-disclosure by influencers, which fails to explain the current phenomena. Here, we focus on negative self-disclosure and explore whether an influencer’s proactive disclosure of failures can enhance purchase intentions, along with the underlying mechanisms and boundary conditions of this strategy. We conducted three online experiments on the Credamo platform. Study 1 (N = 94) explored the main and mediating effect, whereas Study 2 (N = 238) and Study 3 (N = 238) investigated the moderating effects of observer and influencer characteristics, respectively. The following conclusions are drawn: (1) Influencers’ proactive disclosures of failures can boost purchase intentions when recommending products; (2) perceived authenticity plays a mediating role in this process; (3) the degree of viewers’ self-discrepancy moderates the mediating effect of perceived authenticity; and (4) influencers’ follower scale moderates the impact on purchase intentions. This study offers practical implications for influencers on how to enhance marketing effectiveness through self-disclosure. Full article

Energy generation from renewable sources has increased exponentially worldwide, particularly wind energy, which is converted into electricity through wind turbines. The growing demand for renewable energy has driven the development of horizontal-axis wind turbines with larger dimensions, as the energy captured is proportional to the area swept by the rotor blades. In this context, the dynamic loads typically observed in wind turbine towers include vibrations caused by rotating blades at the top of the tower, wind pressure, and earthquakes (less common). In offshore wind farms, wind turbine towers are also subjected to dynamic loads from waves and ocean currents. Vortex-induced vibration can be an undesirable phenomenon, as it may lead to significant adverse effects on wind turbine structures. This study presents a two-dimensional transient model for a rigid body anchored by a torsional spring subjected to a constant velocity flow. We applied a coupling of the Fourier pseudospectral method (FPM) and immersed boundary method (IBM), referred to in this study as IMERSPEC, for a two-dimensional, incompressible, and isothermal flow with constant properties—the FPM to solve the Navier–Stokes equations, and IBM to represent the geometries. Computational simulations, solved at an aspect ratio of $\varphi =4.0$, were analyzed, considering Reynolds numbers ranging from $Re=150$ to Re = 1000 when the cylinder is stationary, and $Re=250$ when the cylinder is in motion. In addition to evaluating vortex shedding and Strouhal number, the study focuses on the characterization of space–time symmetry during the galloping response. The results show a spatial symmetry breaking in the flow patterns, while the oscillatory motion of the rigid body preserves temporal symmetry. The numerical accuracy suggested that the IMERSPEC methodology can effectively solve complex problems. Moreover, the proposed IMERSPEC approach demonstrates notable advantages over conventional techniques, particularly in terms of spectral accuracy, low numerical diffusion, and ease of implementation for moving boundaries. These features make the model especially efficient and suitable for capturing intricate fluid–structure interactions, offering a promising tool for analyzing wind turbine dynamics and other similar systems. Full article

Aircraft wing surface icing compromises flight safety, where accurate calculation of heat transfer coefficient on airfoil surfaces serves as a prerequisite for designing thermal anti-icing systems. However, during icing conditions, ice morphology changes wall roughness and transition properties, making it difficult to accurately determine the heat transfer coefficient. The current study develops a partitioned rough-wall boundary layer integral methodology in order to overcome this issue, extending the conventional boundary layer integral method. The technique generates a convective heat transfer coefficient formulation for aircraft icing surfaces while accounting for roughness differences brought on by water droplet shape. The results show that the partitioned rough-wall boundary layer integral method divides the wing surface into three distinct zones based on water droplet dynamics—a smooth zone, rough zone, and runback zone—each associated with specific roughness values. The NACA0012 airfoil was used for numerical validation, which showed that computational and experimental data concur well. Additionally, the suggested approach predicts transition locations with a high degree of agreement with experimental results. Full article

The lack of new materials with desired processability and functional characteristics remains a challenge for metal additive manufacturing (AM). Therefore, in this work, a new promising AISI 316L-based alloy with better performance compared to the commercially available one is developed via the laser powder bed fusion (L-PBF) process. Moreover, establishing process–structure–properties linkages is a critical point that should be evaluated carefully before adding newly developed alloys into the AM market. Hence, the current study investigates the influences of various process parameters on the as-built quality and microstructure of the newly developed alloy. The results revealed that increasing laser energy density led to reduced porosity and surface roughness, likely due to enhanced melting and solidification. Microstructural analysis revealed a uniform distribution of copper within the austenite phase without forming any agglomeration or secondary phases. Electron backscatter diffraction analysis indicated a strong texture along the build direction with a gradual increase in Goss texture at higher energy densities. Grain boundary regions exhibited higher local misorientation and dislocation density. These findings suggest that changing the process parameters of the L-PBF process is a promising method for developing tailored microstructures and chemical compositions of commercially available AISI 316L stainless steel. Full article

In this paper, a solution to the problem of electromagnetic field scattering on a periodic, constrained, planar antenna structure placed on the boundary of two dielectric media was formulated. The scattering matrix of such a structure was derived, and its generalization for the case of an antenna with a multilayer dielectric substrate was defined. By applying the Galerkin spectral method, the problem was reduced to a system of algebraic equations for the coefficients of current distribution on metal elements of the antenna grid, considering the distribution of the electromagnetic field on Floquet harmonics. The finite transverse dimension of the antenna was considered by introducing, to the solution of the case of an unconstrained antenna, a window function on the antenna aperture. The presented formalism allows modeling the operation of periodic, dielectric, composite antenna arrays. Full article

The current study demonstrates the buckling properties of composite laminates reinforced with MWCNT fillers using a novel higher-order shear and normal deformation theory (HSNDT), which considers the effect of thickness in its mathematical formulation. The hybrid HSNDT combines polynomial and hyperbolic functions that ensure the parabolic shear stress profile and zero shear stress boundary condition at the upper and lower surface of the plate, hence removing the need for a shear correction factor. The plate is made up of carbon fiber bounded together with polymer resin matrix reinforced with MWCNT fibers. The mechanical properties are homogenized by a Halpin–Tsai scheme. The MATLAB R2019a code was developed in-house for a finite element model using C⁰ continuity nine-node Lagrangian isoparametric shape functions. The geometric nonlinear and linear stiffness matrices are derived using the principle of virtual work. The solution of the eigenvalue problem enables estimation of the critical buckling loads. A convergence study was carried out and model efficiency was corroborated with the existing literature. The model contains only seven degrees of freedom, which significantly reduces computation time, facilitating the comprehensive parametric studies for the buckling stability of the plate. Full article

This paper analyzes the transient voltage stability of the dual-source DC power system. The system’s equivalent model is first established. Subsequently, the effect mechanisms of line parameters and voltage-source rectifiers’ current control inner loops on the system’s transient voltage instability are investigated. It indicates that these factors reduce the power supply capacity of the source, increasing the risk of transient instability in the system. Then, considering the influence of fault depths, the influence of different large disturbances on the transient voltage stability is investigated. Furthermore, the critical cutting voltage and critical cutting time for DC power systems are determined and then validated on the MATLAB R2023b/Simulink platform. Finally, based on the mixed potential function theory, the impact of system parameter variations on stability boundaries is analyzed quantitatively. Simulation verification is conducted on the MATLAB R2023b/Simulink platform, and experimental verification is conducted on the RT-LAB Hardware-in-the-Loop platform. The results of the quantitative analysis and experiments corroborate the conclusions drawn from the mechanistic analysis, underscoring the critical role of line parameters and converter control parameters in the system’s transient voltage stability. Full article