Search Results (59,109)

Hyperspectral image (HSI) classification is pivotal in remote sensing, yet deep learning models, particularly Transformers, remain susceptible to spurious spectral–spatial correlations and suffer from limited interpretability. These issues stem from their inability to model the underlying causal structure in high-dimensional data. This paper introduces the Causal Attention Transformer (CAT), a novel architecture that integrates causal inference with a hierarchical CNN-Transformer backbone to address these limitations. CAT incorporates three key modules: (1) a Causal Attention Mechanism that enforces temporal and spatial causality via triangular masking and axial decomposition to eliminate spurious dependencies; (2) a Dual-Path Hierarchical Fusion module that adaptively integrates spectral and spatial causal features using learnable gating; and (3) a Linearized Causal Attention module that reduces the computational complexity from $O\left(N^2\right)$ to $O\left(N\right)$ via kernelized cumulative summation, enabling scalable high-resolution HSI processing. Extensive experiments on three benchmark datasets (Indian Pines, Pavia University, Houston2013) demonstrate that CAT achieves state-of-the-art performance, outperforming leading CNN and Transformer models in both accuracy and robustness. Furthermore, CAT provides inherently interpretable spectral–spatial causal maps, offering valuable insights for reliable remote sensing analysis. Full article

Ultrasonic-assisted laser-directed energy deposition (UA-DED) is a promising combined technology for manufacturing high-value thin-walled Inconel 718 components in aerospace. Nevertheless, the optimal ultrasonic frequency—a key parameter for achieving desirable performance in thin-walled Inconel 718 alloys—remains to be determined. In this study, we systematically investigated the influence of ultrasonic frequency (12–20 kHz) on the microstructure and mechanical properties of thin-walled Inconel 718 fabricated by UA-DED. The results revealed that an ultrasonic frequency of 20 kHz was optimal and can yield significant improvements in the microstructures of the as-deposited sample coordinate planes, manifested by the complete suppression of large pores, three-dimensional refinement of the γ matrix grains, alleviation of Nb and Mo segregation, the reduction of fragmented Laves particles, a decrease in residual macroscopic stresses, and homogeneous distributions of γ′/γ″ phases and γ-grain orientation. Meanwhile, the application of a 20 kHz ultrasonic frequency endows the manufactured thin-walled 718 parts with superior mechanical properties, including a tensile strength of 899 MPa in the laser scanning direction and 877 MPa in the build direction, along with the corresponding elongations of 34.8% and 38.9%. This work demonstrates the potential of modulating ultrasonic frequency to tailor microstructures and produce high-performance thin-walled Inconel 718 aerospace components. Full article

A systematic density functional theory (DFT) and time-dependent DFT (TD-DFT) investigation of aluminum phosphide (Al_xP_x) nanoparticles with diverse dimensionalities and geometries is presented. Starting from a cubic-like Al₄P₄ building block, a series of one-dimensional (1D) elongated, two-dimensional (2D) exotic, and extended sheet-like nanostructures was constructed, enabling a unified structure–property analysis across size and topology. Optical absorption and infrared (IR) vibrational spectra were computed and correlated with geometric motifs, revealing pronounced shape-dependent tunability. Compact and highly interconnected 2D architectures exhibit red-shifted absorption and enhanced vibrational polarizability, whereas elongated or low-connectivity motifs lead to blue-shifted optical responses and stiffer vibrational frameworks. Benchmark comparisons indicate that CAM-B3LYP excitation energies closely reproduce reference EOM-CCSD trends for the lowest singlet states. Binding energy and HOMO-UMO gap analyses confirm increasing thermodynamic stability with size and dimensionality, alongside topology-driven electronic modulation. These findings establish AlP nanostructures as highly tunable platforms for optoelectronic and vibrationally active applications. Full article

To address the urgent need for high-fidelity motion data for validating navigation algorithms for Unmanned Underwater Vehicles (UUVs), this paper proposes a data generation method based on a parametric motion model. First, based on the principles of rigid body dynamics and fluid mechanics, a decoupled six-degrees-of-freedom (6-DOF) Linear and Angular Acceleration Vector (LAAV) model is constructed, establishing a dynamic mapping relationship between the rudder angle and speed setting commands and motion acceleration. Second, a segmentation–identification framework is proposed for three-dimensional trajectory segmentation, integrating Gaussian Process Regression and Ordering Points To Identify the Clustering Structure (GPR-OPTICS), along with a Dynamic Immune Genetic Algorithm (DIGA). This framework utilizes real vessel data to achieve motion segment clustering and parameter identification, completing the construction of the LAAV model. On this basis, by introducing sensor error models, highly credible Inertial Measurement Unit (IMU) data are generated, and a complete attitude, velocity, and position (AVP) motion sequence is obtained through an inertial navigation solution. Experiments demonstrate that the AVP data generated by our method achieve over 88% reliability compared with the real vessel dataset. Furthermore, the proposed method outperforms the PSINS toolbox in both the reliability and accuracy of all motion parameters. These results validate the effectiveness and superiority of our proposed method, which provides a high-fidelity data benchmark for research on underwater navigation algorithms. Full article

The gut microbiome is a modifiable factor in cancer survivorship. Diet represents the most practical intervention for modulating the gut microbiome. However, diet–microbiome relationships in prostate-cancer survivors remain poorly characterized. We conducted a comprehensive analysis of diet–microbiome associations in 79 prostate-cancer survivors (ages 62–81) enrolled in a randomized exercise intervention trial, 59.5% of whom still have active metastatic disease. Dietary intake was assessed using the Diet History Questionnaire (201 variables) and analyzed using three validated dietary pattern scores: Mediterranean Diet Adherence Score (MEDAS), Healthy Eating Index-2015 (HEI-2015), and the Mediterranean-Dash Intervention for Neurodegenerative Delay (MIND) diet score. Gut microbiome composition was characterized via 16S rRNA sequencing. Dimensionality reduction strategies, including theory-driven diet scores and data-driven machine learning (Random Forest, and Least Absolute Shrinkage and Selection Operator (LASSO)), were used. Statistical analyses included beta regression for alpha diversity, Permutational Multivariate Analysis of Variance (PERMANOVA) for beta diversity (both Bray–Curtis and Sørensen metrics), and Microbiome Multivariable Associations with Linear Models (MaAsLin2) with negative binomial regression for taxa-level associations. All models tested interactions with exercise intervention, APOLIPOPROTEIN E (APOE) genotype, and testosterone levels. There was an interaction between MEDAS and exercise type on gut alpha diversity (Shannon: p = 0.0022), with stronger diet–diversity associations in strength training and Tai Chi groups than flexibility controls. All three diet-quality scores predicted beta diversity (HEI p = 0.002; MIND p = 0.025; MEDAS p = 0.034) but not Bray–Curtis (abundance-weighted) distance, suggesting diet shapes community membership rather than relative abundances. Taxa-level analysis revealed 129 genera with diet associations or diet × host factor interactions. Among 297 dietary variables tested for cognitive outcomes, only caffeine significantly predicted Montreal Cognitive Assessment (MoCA) scores after False Discovery Rate (FDR) correction (p = 0.0009, q = 0.014) through direct pathways beneficial to cognitive performance without notable gut microbiome modulation. In cancer survivors, dietary recommendations should be tailored to exercise habits, genetic background, and hormonal status. Full article

In a situation where cities are increasingly being developed vertically and complexly, a novel approach for analyzing vertical urban functional patterns is proposed. For this purpose, a multi-dimensional GDC (Geo Data Cube) consisting of spatial and temporal data x, y, z, t, and f dimensions containing layer information was created. At this time, the size of the GDC cell (interval in x, y, z dimensions) is calculated by cell point data using the three-dimensional (3D) Moran’s I index value calculated with the 3D Diversity Factor (DF) based on information entropy proposed to reduce the uncertainty of information for each cell. In other words, the cell with the smallest index value was chosen to minimize the influence of Modifiable Areal Unit Problem (MAUP) that occurs when mapping. The 3D land use index (3D LUI) is calculated as a linearly weighted sum of the spatial accessibility of uses between cells (3D KDF) and the enrichment of uses (3D EF), taking into account the first law of geography. Finally, the 3D LUI value for each use was calculated for each cell of the GDC, and the use with the highest value was determined as the urban function of the cell. As a result of applying this to Seocho-gu, Seoul, Republic of Korea (ROK) in June 2024 and visually evaluating it using the street view provided by Kakao Map, it was confirmed that commercial and residential functions were vertically separated in buildings with residential–commercial complexes or shops on the ground floor. It was also confirmed that such characteristics did not appear in the two-dimensional (2D) urban functional patter analysis. Full article

This study evaluates the effects of hydrothermal and oleothermal treatments on the physical, colorimetric, and mechanical properties of Dabema wood. Samples were heated at 100, 160, and 220 °C for 2, 3.5, and 5 h. Equilibrium moisture content decreased from 13.16% in untreated wood to approximately 43% lower after hydrothermal treatment at 160 °C for 5 h and to 64% lower after oleothermal treatment at 220 °C for 5 h. Water absorption decreased from 78% in untreated samples to 25%–64% following hydrothermal treatment and to 17%–44% after oleothermal treatment. Hydrothermal treatment caused significant darkening, whereas oleothermal treatment maintained a lighter, more stable color. Mechanical properties improved substantially: in compression, MOE increased by 113% after oleothermal treatment at 220 °C for 5 h. In bending, MOR and MOE rose by 25%–35% under optimal oil-heat conditions. In tensile, MOE increased by 30%, and maximum tensile stress improved by up to 130%. Oleothermal treatments yielded the most stable enhancements, whereas severe hydrothermal treatments sometimes reduced mechanical performance despite improving moisture resistance. Multivariate analysis (PCA) and response surface methodology (RSM) indicate that oleothermal treatment at 160 °C for 3.5–5 h provides the best compromise between stiffness and color stability. Thermogravimetric analyses (TG/DTG) show hydrothermal treatment promotes hemicelluloses degradation, whereas oleothermal treatment stabilizes the cellulose–lignin network. Overall, hydrothermal treatment enhances dimensional stability, while oleothermal treatment achieves an optimal balance of stiffness, mechanical performance, and color retention. Deep color changes from furanic resin formation under hydrothermal conditions are strongly suppressed by oil during oleothermal processing, yielding lighter and more durable wood. For commercial applications such as furniture and structural components, oleothermal treatment is recommended, whereas hydrothermal treatment is more suitable when dimensional stability is prioritized over mechanical performance. Full article

Suffusion is a primary cause of failure in hydraulic structures, including earth dams; however, the mechanisms underlying suffusion-induced failure and the stability changes remain poorly understood. This study derives and implements a sequentially coupled computational model that considers the effect of seepage–suffusion–stress, aimed at simulating the entire process of suffusion-induced failure in earth dams and evaluating their stability. The accuracy of the proposed approach is validated through comparisons with one-dimensional consolidation theory, suffusion experiments, and triaxial tests on eroded soil. A model of the earth dam at high water levels is developed to simulate the full process of suffusion-induced failure and assess its stability. The results indicate that, under the influence of suffusion, fines are lost most rapidly at the dam toe, followed by the region near the upstream water level. In the later stages of suffusion, the soil near the slip surface undergoes excessive compression, leading to an increase in fine content rather than a decrease. The mechanism of suffusion-induced failure in earth dams involves severe fines loss at the dam toe and near the upstream water level, which leads to significant soil weakening and the formation of a continuous plastic zone extending from the dam toe to the upstream water level. The safety factor of the earth dam, when suffusion effects are not considered, remains nearly constant, making it challenging to accurately assess its stability. The safety factor of the earth dam remains nearly constant when suffusion is neglected, indicating that overlooking suffusion presents substantial safety risks. Furthermore, reducing the permeability coefficient of the earth dam can effectively mitigate suffusion. Full article

Background/Objectives: The diffuse sclerosing variant of papillary thyroid carcinoma (DSV-PTC) is a rare and aggressive subtype characterized by diffuse gland involvement and early cervical lymph node metastasis. Preoperative differentiation from classic papillary thyroid carcinoma and autoimmune thyroid disease remains challenging on B-mode ultrasound. This study aimed to describe the multiparametric ultrasound features of DSV-PTC in a single-center case series and highlight practical imaging insights. Methods: We retrospectively reviewed seven consecutive patients with histologically confirmed DSV-PTC evaluated at a single center between 2013 and 2025. All patients underwent standardized B-mode ultrasound, color Doppler, and two-dimensional shear-wave elastography prior to surgery. Clinical, autoimmune, cytological, surgical, pathological, and follow-up data were analyzed descriptively. Results: The cohort included five females and two males (mean age 28 years). Autoimmune thyroid disease was present in three patients. High-risk ultrasound features were identified in all cases, with microcalcifications in six patients and a diffuse “snowstorm” appearance in five. Elastography demonstrated increased stiffness in six out of seven lesions (Emean 28–173 kPa; Emax 31–300 kPa). Cervical lymph node metastases were confirmed in all patients. In two cases, elastography aided identification of focal malignant involvement within diffusely altered thyroid parenchyma. All patients underwent total thyroidectomy with central neck dissection; lateral neck dissection and radioiodine therapy were performed selectively. No distant metastases were detected. Conclusions: In this case series, DSV-PTC showed a characteristic multiparametric ultrasound pattern combining high-risk B-mode features with frequently increased tissue stiffness. Elastography provided complementary information, particularly in the presence of autoimmune thyroid disease, by helping localize focal malignant involvement within diffusely altered parenchyma. Full article

Background/Objectives: Optimizing routine neurointerventional workflow and minimizing exposure to ionizing radiation during coil-only endovascular treatment of intracranial aneurysms depend on operator experience, reduced frame rates during both fluoroscopy and digital subtraction angiography (DSA), and the use of advanced angiographic systems. The low-dose protocol implemented in this study used the lowest available fluoroscopy frame rate (3.125 frames per second [fps]) and a nominal acquisition rate of 2 fps (actual = 2.45 fps) for DSA, three-dimensional (3D) rotational angiography, two-dimensional (2D)/3D mapping, and roadmapping. Methods: This retrospective analysis encompassed 245 coil-only procedures performed at a single tertiary center from 2018 to 2024. Data collected for each procedure included dose-area product (DAP), reference air kerma (K_a,r), fluoroscopy time (FT), and the total number of DSA frames. Local diagnostic reference levels (DRLs; 75th percentile [P75]) and typical values (50th percentile [P50]) were determined and descriptively compared with values reported in the literature. Results: The P75 values, representing DRLs, were 22.4 Gy·cm² for DAP (literature range, 123–272.8 Gy·cm²), 268 mGy for K_a,r (1171–4240 mGy), 18 min 56 s for FT, and 285 DSA frames. The P50 values were 13.8 Gy·cm² for DAP (78.7–179.0 Gy·cm²), 196 mGy for K_a,r (801–2804 mGy), 13 min 25 s for FT, and 208 DSA frames. Conclusions: In this single-center cohort, dose metrics for coil-only intracranial aneurysm treatment were within the lower range of published values. Cross-study comparisons are descriptive and require cautious interpretation. The proposed local DRLs may support quality assurance, dose optimization, and patient safety in comparable clinical settings. Further multi-center and multi-operator studies are warranted to evaluate transferability and applicability beyond coil-only procedures. Full article

Fault diagnosis is a core technology ensuring the safe and efficient operation of industrial systems. A paradigm shift has been observed wherein traditional signal analysis has been replaced by intelligent, algorithm-driven approaches. In recent years, large language models, digital twins, and knowledge graphs have been introduced. A new stage of intelligent integration has been reached that is characterized by data-driven methods, knowledge guidance, and physical–virtual fusion. In the present paper, the evolutionary context of fault diagnosis technologies was systematically reviewed, with a focus on the theoretical methods and application practices of traditional machine learning, digital twins, knowledge graphs, and large language models. First, the research background, core objectives, and development history of fault diagnosis were described. Second, the principles, industrial applications, and limitations of supervised and unsupervised learning were analyzed. Third, innovative uses were examined involving physical–virtual mapping in digital twins, knowledge modeling in knowledge graphs, and feature learning in large language models. Subsequently, a multi-dimensional comparison framework was constructed to analyze the performance indicators, applicable scenarios, and collaborative potential of different technologies. Finally, the key challenges faced in the current fault diagnosis field were summarized. These included data quality, model generalization, and knowledge reuse. Future directions driven by the fusion of large language models, digital twins, and knowledge graphs were also outlined. A comprehensive technical map was established for fault diagnosis researchers, as well as an up-to-date reference. Theoretical innovation and engineering deployment of intelligent fault diagnosis are intended to be supported. Full article

Multi-dimensional inequality among rural residents has become a major obstacle hindering the achievement of global poverty alleviation goals. This study utilized household sample data from the China Family Panel Studies (CFPS) over four periods from 2014 to 2020 and applied them to a high-dimensional fixed effects model to estimate the impact of digital literacy on multi-dimensional inequality among rural residents. The results show that digital literacy can effectively alleviate the multi-dimensional inequality of rural residents. From the perspective of a mediating effect, digital literacy alleviates the multi-dimensional inequality of rural residents by improving the level of social capital and promoting social harmony. Moreover, the alleviation of multi-dimensional inequality among rural residents by digital literacy varies among different groups. The impact of digital literacy on the multi-dimensional inequality of agricultural workers and rural residents in western regions is relatively greater than that of non-agricultural workers and rural residents in other regions. Information processing literacy in digital literacy has the most significant impact on the multi-dimensional inequality of rural residents. This paper enriches the mechanism paths of digital literacy in alleviating the multi-dimensional inequality among rural residents in terms of both material and spiritual aspects, and provides a certain reference value for achieving the all-round development of rural residents and contributing to rural production practices. Full article

As vital components of urban rail transit networks, subway stations are widely scattered across diverse urban districts, whose sustainability performance exerts a notable impact on the overall urban ecological and environmental quality. This study constructs a three-dimensional numerical model to conduct a comparative assessment of the seismic behavior of subway stations adopting different bearing systems at beam-column joints. The seismic responses of two typical structural configurations, a traditional rigid-jointed subway station and another equipped with rubber isolation bearings, are examined under a series of ground motions, with due consideration of amplitude scaling effects and material nonlinearity. A comprehensive evaluation is carried out on key performance parameters, including structural acceleration responses, column rotation angles, damage evolution processes, and internal force distributions. Based on this analysis, the research clarifies the sustainability implications by establishing quantitative correlations between seismic response indices (i.e., deformation extent, damage degree, and internal force magnitudes) and post-earthquake outcomes, such as repair complexity, material requirements, carbon emissions, and socioeconomic effects. The results can advance the integrated theory of seismic-resilient and sustainable design for underground infrastructure, providing evidence-based guidance for the optimization of future subway station construction projects. Full article

Automatic sleep-stage classification is a key component of long-term sleep monitoring and digital health applications. Although deep learning models trained on centralized datasets have achieved strong performance, their deployment in real-world healthcare settings is constrained by privacy, data-governance, and regulatory requirements. Federated learning (FL) addresses these issues by enabling decentralized training in which raw data remain local and only model parameters are exchanged; however, its effectiveness under realistic physiological heterogeneity remains insufficiently understood. In this work, we investigate a subject-level federated deep learning framework for sleep-stage classification using polysomnography data from the ISRUC-Sleep dataset. We adopt a realistic one subject = one client setting spanning three clinically distinct subgroups and evaluate a lightweight one-dimensional convolutional neural network (1D-CNN) under four training regimes: a centralized baseline and three federated strategies (FedAvg, FedProx, and FedBN), all sharing identical architecture and preprocessing. The centralized model, trained on a cohort with regular sleep architecture, achieves stable performance (accuracy 69.65%, macro-F1 0.6537). In contrast, naive FedAvg fails to converge under subject-level non-IID data (accuracy 14.21%, macro-F1 0.0601), with minority stages such as N1 and REM largely lost. FedProx yields only marginal improvement, while FedBN—by preserving client-specific batch-normalization statistics—achieves the best federated performance (accuracy 26.04%, macro-F1 0.1732) and greater stability across clients. These findings indicate that the main limitation of FL for sleep staging lies in physiological heterogeneity rather than model capacity, highlighting the need for heterogeneity-aware strategies in privacy-preserving sleep analytics. Full article

Background: Urinary proteins may reflect physiological changes occurring during the periparturient period, but reference data for goats are still lacking. This study investigated urinary protein patterns around parturition to help fill this gap and generate baseline information. Methods: Ten pregnant Alpine goats were sampled by spontaneous voiding 22 ± 3 days before delivery (T0), 7 days postpartum (T7), and 30 days postpartum (T30). Physical and chemical urine analyses were performed, and urinary proteins were separated using one-dimensional sodium dodecyl sulfate–polyacrylamide gel electrophoresis. Statistical tests (Shapiro–Wilk, repeated-measures ANOVA, or Friedman) were applied to evaluate differences among time points. Results: Significant temporal changes were observed: urine pH decreased at T30, the urine protein-to-creatinine ratio increased at T7 and T30, and urinary creatinine concentration was highest at T0. Most samples showed common protein bands at approximately 80, 70, 62, 50, 37, 29, 25, 22, and below 13 kDa, with the 62, 50, and <13 kDa bands present in all samples. Bands between 18 and 64 kDa and above 60 kDa appeared only in some samples. Protein bands between 23 and 42 kDa were more frequent at T0, suggesting immune-related variations associated with pregnancy. Conclusions: This study provides the first description of urinary protein electrophoretic profiles in goats during the periparturient period and highlights measurable changes across time. These findings offer a starting point for developing future research and may contribute to establishing reference parameters for clinical and physiological monitoring in goats. Full article