Search Results (37,944)

This manuscript presents the theory and experimental validation of low- and medium-temperature deicing technology that advances the recently developed Ice Cavitation Deicing (ICD) method. Conventional high-temperature ICD (HTICD) efficiently removes ice by explosively vaporizing a thin interfacial melted layer but operates at heating rates above 10⁶ K/s, high voltage, and maximum temperatures exceeding 400 °C. This study develops Low-Temperature and Medium-Temperature Ice Cavitation Deicing (LTICD and MTICD), extending ICD into the previously unexplored intermediate heating-rate regime. Analytical modeling based on energy conservation, transient heat diffusion, water thermodynamics, and thermal-stress analysis was combined with finite-element simulations and experimental testing. Several foil materials were evaluated over heating rates of approximately 10⁴–10⁷ K/s using capacitor banks of 0.1–35 mF. Experiments demonstrated effective removal of thick and thin ice at cavitation temperatures of approximately 120–200 °C, substantially below those of HTICD. The lower operating temperatures and heating rates reduce thermal stress, voltage, and current, enable practical low-voltage electrolytic capacitors, and expand the range of suitable materials. Thus, LTICD and MTICD provide a lower-temperature, more practical electrical architecture for future aircraft ice-protection systems. Full article

Automated prostate cancer (PCa) lesion segmentation using deep learning remains constrained by limited tissue contrast in standard diffusion-based MRI sequences, with state-of-the-art methods reporting Dice scores of 32% or lower on large patient cohorts. Synthetic correlated diffusion imaging (${\mathrm{CDI}}^s$) offers a promising solution, providing enhanced tissue contrast derived entirely from existing diffusion-weighted imaging (DWI) acquisitions at no additional clinical cost. This study presents the first comprehensive evaluation of ${\mathrm{CDI}}^s$ integration across the full standard multiparametric MRI protocol, encompassing 15 modality configurations and six segmentation architectures spanning CNN and transformer families on a cohort of 200 patients. ${\mathrm{CDI}}^s$ reliably enhances or preserves segmentation performance in the evaluated configurations, with 19 statistically significant improvements and no significant degradations across 42 direct comparisons. ${\mathrm{CDI}}^s$ enhancement primarily operates as a recall-driven mechanism, improving lesion detection sensitivity while largely preserving precision. ${\mathrm{CDI}}^s$ + DWI + T2w emerged as the strongest clinically meaningful configuration, achieving significant Dice improvement in four of six architectures, with no instances of degradation. Grad-CAM-based explainability analysis further reveals that ${\mathrm{CDI}}^s$ focuses poorly localized CNN attention toward lesion boundaries, while transformer architectures exhibit more stable attention patterns that are less sensitive to ${\mathrm{CDI}}^s$ integration. These results establish validated ${\mathrm{CDI}}^s$ integration pathways and provide architecture-specific deployment guidance for clinical implementation. Full article

The primary cause of the scale effect in scaled models for flood discharge and energy dissipation lies in the dissimilarity of the air dispersion patterns of the ski-jump nappe. To uncover the scale-effect relationship governing the air dispersion patterns of ski-jump energy dissipation nappes in high dams, a series of scaled physical model tests were conducted at the Baihetan Hydropower Station. The air dispersion patterns were systematically observed, as well as the distribution characteristics of entrained air concentration in the ski-jump nappe across various scales. Based on the experimental observations, a two-dimensional stochastic diffusion numerical model was developed, successfully replicating the dispersion process of the nappe as it gradually transformed from a crescent shape to a circular one. Furthermore, by calibrating the concentration distribution curve, a quantitative relationship was established between the random displacement parameter σ and the Weber number. The study revealed that when the Weber number (We) is below 40,000, σ increases rapidly and approximately linearly with We, indicating a high sensitivity to dispersion degree. However, once We surpasses 40,000, the growth rate significantly decelerates, approaching saturation, suggesting that the dispersion degree closely approximates the prototype condition. Consequently, it is suggested that the Weber number control threshold for the physical model of ski-jump water–air two-phase flow in high dams be set above 40,000, providing a valuable reference for selecting large-scale models. Full article

Generative artificial intelligence (GenAI) has diffused through higher education faster than institutions have been able to govern it, reshaping the conditions under which universities produce knowledge, judgement, credentials, and public trust. Current responses (prohibition, detection, and accommodation) fall short of a settled governance posture, and existing frameworks, from AI ethics principles to standard Responsible Research and Innovation (RRI) models, are not calibrated to higher education’s distinctive epistemic, formative, and public-good missions. This article addresses that gap through a disciplined conceptual synthesis drawing on RRI, reflexive governance, and higher education theory. The synthesis develops a Reflexive and Responsible Innovation Paradigm (RRIP): a six-dimensional framework that re-specifies RRI’s canonical dimensions (anticipation, reflexivity, inclusion, responsiveness) for the university context and adds two higher-education-specific dimensions: epistemic stewardship and distributive justice. Epistemic stewardship, the article’s central theoretical contribution, names the institutional obligation to protect the conditions under which knowledge claims are formed, warranted, assessed, and trusted under AI mediation. RRIP is operationalized through a multi-level architecture of institutional mechanisms (deliberative AI councils, transparency registers, and reflexive assessment redesign) with a tiered implementation pathway calibrated to institutions of varying capacity. Institutional leaders, program directors, policymakers, and accreditation bodies will find in RRIP a theoretically grounded and practically applicable guide for assessment redesign, curriculum decisions, procurement governance, and sectoral coordination. Full article

Molecular transport of low-volatility organic vapors through unsaturated porous media is governed by the coupled effects of diffusion, sorption, and pore structure, yet the relative roles of these processes remain insufficiently quantified. In this work, we develop a physics-based model describing vapor migration from a subsurface source to the soil surface by explicitly accounting for moisture-dependent sorption, air-filled porosity, and pore clogging by fine particles. The model predicts that soil moisture affects vapor transport through two competing mechanisms: thin water films progressively suppress gas–solid sorption, thereby increasing the effective diffusion coefficient, whereas further wetting reduces the connectivity of air-filled pores and ultimately blocks gas-phase transport. As a consequence, vapor migration exhibits a non-monotonic dependence on soil moisture, with a distinct optimum for surface vapor flux. The model further predicts that fine particles substantially decrease vapor transport by reducing pore connectivity, while lower temperatures suppress migration through both reduced molecular diffusivity and enhanced sorption. Laboratory experiments using representative low-volatility energetic compounds through sand with controlled moisture content and particle composition confirmed all major qualitative predictions of the model, including enhanced transport at intermediate moisture, suppression under dry, dusty and highly saturated conditions, and strong temperature dependence. Although energetic compounds were used as representative low-volatility substances, the proposed framework is generally applicable to molecular transport of trace organic vapors in unsaturated porous media and provides a quantitative basis for predicting environmental conditions under which subsurface sources can be detected by surface vapor measurements. Full article

Low-light image super-resolution aims to recover normal-light high-resolution images from dark low-resolution observations captured by image sensors, in which illumination attenuation, sensor noise, blur, and low resolution are entangled, making it more challenging than conventional super-resolution. Diffusion-based methods perform well on real-world super-resolution but usually require costly multi-step inference; recent one-step methods either rely on a globally fixed timestep that cannot adapt to per-sample degradation, or they directly encode the dark image into the latent space, coupling illumination bias with content degradation. To address these issues, we propose RASR, a Retinex-guided adaptive one-step diffusion framework for low-light super-resolution. We first decompose the observation into reflectance and illumination, and we use the reflectance as the content carrier to align it with the normal-light prior of the pretrained model. A latent-space teacher then constructs per-sample supervision from the low/high-quality latent discrepancy, while a lightweight student predicts the noise level solely from the Retinex representation, removing the dependence on high-quality references at inference. Finally, a single velocity-field integration on Stable Diffusion 3 yields the result, updating only low-rank adapters and lightweight modules during training. Extensive experiments on the RELLISUR benchmark show that RASR overall outperforms existing low-light and mainstream super-resolution methods in PSNR, SSIM, and LPIPS, with especially prominent gains in perceptual quality, and ablation studies validate the effectiveness of each key design. Full article

Countries operating major container ports without a domestic equipment supply base are pursuing self-reliance in smart port technology, but such programmes are hard to assess because their effects diffuse across the economy through inter-industry linkages. This study estimates inducement coefficients for the Korean Smart Port Technology Self-Reliance Equipment Development Project. Using the 2023 Input–Output Tables of the Bank of Korea, the programme is defined at the basic-sector level: 67 of the 380 basic sectors, spanning port equipment manufacture, construction, logistics services, software, and R&D, are extracted from their parent groups and consolidated into one sector, leaving the unrelated residuals endogenous. That sector is exogenously specified, capturing only repercussions on the remaining 33 sectors. A KRW 1 increase in programme output induces KRW 0.8358 of production and KRW 0.2958 of value added elsewhere, and KRW 1 billion induces 2.9758 jobs; on the domestic table these fall to KRW 0.5163, KRW 0.1872, and 1.9912 jobs, so about two-fifths of the gross inducement leaks abroad through imports. Production inducement falls upstream in materials, value-added, and employment inducement downstream in services. On the supply side, a KRW 1 shortfall in the sector’s domestic supply disrupts KRW 0.4013 of production among its users, and the sector shows the highest forward-linkage sensitivity of the 34-sector system. Eleven alternative delineations leave the structural findings unchanged. The coefficient vector is reported in full, so the estimates are reproducible. Full article

Fused silica–aluminum alloy dissimilar connections are in urgent demand in fields such as aerospace optoelectronic packaging, vacuum optical windows, and micro-electro-mechanical systems, yet the dramatic mismatch in thermal-expansion coefficient and thermophysical properties between the two materials has long been a bottleneck for reliable joining. Current ultrafast laser welding of such heterogeneous systems still suffers from prominent problems, including stringent optical contact requirements, high crack sensitivity on the fused silica side, and unclear coupling mechanism between clamping conditions and joint defects. This work systematically studies the joining process of femtosecond laser welding of fused silica and 6061 aluminum alloy dissimilar materials, focusing on the effects of scanning speed, pulse energy, scanning spacing, and fixture preload on the shear strength, microstructure, and elemental diffusion behavior of the joints. The results confirm that scanning speed and scanning spacing have a synergistic effect on heat input density; the magnitude of the fixture preload is a key factor determining the interfacial residual stress and crack sensitivity. By optimizing the scanning speed (6 mm/s) and combining it with a low preload and 140 μm scanning spacing, a high-strength heterogeneous joint with uniform elemental transition and no macroscopic cracks can be obtained. This study provides a detailed process-optimization approach for high-quality laser welding of dissimilar brittle/ductile materials. Full article

This work evaluates agarose (AG)- and chitosan (CS)-based hydrogels incorporating a dexamethasone–cyclodextrin complex (Dex) as potential bioink candidates for wound healing applications. For this purpose, biopolymer hydrogels were formulated and tested to determine their structure, morphology, and rheological behavior under different strain, frequency, and temperature conditions, as well as their swelling behavior, surface functionality, protein adhesion capability, and antibacterial activity. Scanning electron microscopy revealed an interconnected membrane-like porous structure in the agarose/chitosan (BAC) hydrogels, which also showed a fibril-like microstructure when Dex was added. X-ray diffraction showed that these structures had a semicrystalline nature, with a main broad peak from approximately 15 to 20°. Rheological tests confirmed the viscoelastic behavior of the developed hydrogels. Chitosan incorporation reduced the rigidity of the formulations, whereas dexamethasone did not significantly affect the viscoelastic response of the BAC hydrogel. Additionally, the temperature sweep showed that Dex did not substantially alter the viscoelastic response of the BAC hydrogel while reducing mass loss after immersion in phosphate-buffered saline and not significantly affecting the swelling behavior. Further, Coomassie blue staining tests demonstrated favorable protein interaction with the BAC-based hydrogels, indicating that Dex incorporation did not compromise their adsorption capacity. Regarding their antibacterial activity, the hydrogels showed no major effects on Escherichia coli and Staphylococcus aureus in disk diffusion tests. However, liquid diffusion tests showed almost no bacterial growth for hydrogels incorporating CS. Thus, the present results demonstrate the potential of the BAC hydrogel system for biomedical and cosmetic applications. Full article

Constructed wetlands rely on microscale redox heterogeneity to couple ammonium oxidation with downstream nitrogen loss, yet the causal role of radial oxygen loss (ROL) is often inferred from planting effects, bulk redox potential, functional genes, or outlet performance rather than measured directly. This narrative mechanistic review aims to determine when ROL-generated root oxygen microzones become functionally relevant, how they alter competition and coupling among nitrification, denitrification, anammox, DNRA, and NO/N₂O branch points, and whether present evidence supports scaling from individual roots to bed-scale total-nitrogen (TN) performance. We conduct a structured narrative evidence synthesis that separates direct root-zone O₂/ROL measurements from redox-resolved contextual evidence and mechanistic analogy, and appraises studies by measurement directness, resolution, nitrogen-endpoint resolution, operating-context reporting, control of competing explanations, and scale relevance. The broader mechanistic corpus is extensive, but only a small subset contains spatially resolved root-zone oxygen measurements coupled with nitrogen-related endpoints. Direct measurements show strong heterogeneity: reported lateral-root DO is 0.64–2.04 mg L⁻¹ with 0.76–1.16 mm oxygen layers, root-surface oxygen declines from about 2.0 to 0.5 mg L⁻¹ with depth in one Acorus system, and potential radial oxygen diffusion in Typha middle-root sections spans 0.003–0.316 µmol O₂ L⁻¹ µm⁻¹. We therefore treat ROL as a context-dependent oxygen budget rather than a fixed species property. Root-associated oxygen gradients are demonstrated and strongly heterogeneous; their effects on pathway feasibility and spatial coupling are conditionally supported; however, a transferable ROL coefficient that predicts bed-scale TN removal is not yet demonstrated. Progress requires direct root-zone mapping, pathway-rate measurements, complete nitrogen mass balances, internally constrained reactive-transport models, and seasonal field validation. Full article

Background: Differentiating alcohol-related liver disease (ARLD) from chronic liver injury caused by other etiologies remains a clinically relevant challenge in cross-sectional imaging. In particular, alcoholic hepatitis and alcoholic cirrhosis may overlap morphologically with HBV-related cirrhosis on CT imaging. To develop and assess an attention-residual hybrid Convolutional Neural Network (CNN) for multiclass CT image classification of alcoholic hepatitis, alcoholic cirrhosis, HBV-related cirrhosis, and living liver donors. Methods: A four-class liver image dataset comprising 5760 CT images from 144 individuals (36 per group) was constructed using images from alcoholic hepatitis, alcoholic cirrhosis, HBV-related cirrhosis, and living liver donors. The dataset was partitioned into training, validation, and test sets at the patient level. Five pretrained CNN architectures, including DenseNet121, ResNet50, MobileNetV3-Large, EfficientNetB0, and ConvNeXt-Tiny, were first fine-tuned and comparatively evaluated. Based on F1-score ranking, DenseNet121 and ConvNeXt-Tiny were selected as the two backbone networks for the proposed hybrid model. The final architecture integrated Attention Pooling, Feature-wise Linear Modulation (FiLM), Multi-head Attention, Gated Linear Units, Residual Connections, and layer normalization to improve feature fusion and contextual representation. Results: The proposed model demonstrated the best overall performance among all evaluated architectures on the test set. It achieved an accuracy of 99.55%, a weighted F1-score of 99.55%, an MCC of 0.9941, and a Cohen’s kappa coefficient of 0.9940. The model also achieved ROC-AUC and PR-AUC values of 100.00% and 99.99%, respectively, together with an NPV of 99.85%. The proposed model’s performance was also balanced across classes. For Alcoholic Cirrhosis, precision, recall, and F1-score were all 99.64%. For Alcoholic Hepatitis, the corresponding values were 100.00%, 98.93%, and 99.46%, respectively, while HBV-related Cirrhosis achieved 99.29% precision, 99.64% recall, and 99.47% F1-score. Living Liver Donors achieved 99.29% precision, 100.00% recall, and 99.64% F1-score. Conclusions: The findings of this exploratory study suggest that routine CT images may contain image-based differences potentially relevant to etiology-oriented classification of diffuse liver disease. Beyond distinguishing ARLD from HBV-related cirrhosis and images from living liver donors, the model also captured image-based differences between major ARLD subgroups, including alcoholic hepatitis and alcoholic cirrhosis. These findings support further investigation of CT-derived image-based differences in larger independent and multicenter datasets. Full article

:Hierarchical porous catalysts now benefit from intentional co-design of transport pathways and catalytic functionality. This review critically examines intelligent micro–mesoporous nanoarchitectures, emphasizing pore connectivity and active-site cooperativity as inseparable design principles. We first outline limitations of purely microporous systems (diffusion constraints, site inaccessibility, deactivation) and then show how multi-scale networks overcome these issues. Engineering strategies for pore connectivity involving bottom-up templating, post-synthetic reconstruction, top-down desilication/dealumination are systematically reviewed alongside metrics (tortuosity, connectivity, accessibility). Active-site cooperativity is examined via acid-based bifunctionality, metal-acid coupling, single-atom catalysis and compartmentalized architectures for cascade reactions. The central thesis is that optimal performance emerges when transport and catalytic site engineering are coupled, supported by evidence from zeolites, metal–organic frameworks, silica nanoreactors, heteroatom-doped carbons and advanced electrocatalysts. Applications include biomass upgrading, selective oxidation, and energy conversion. The review also covers stability, deactivation, and regeneration, suggests standardized reporting criteria, and highlights future challenges such as using AI for catalyst design, operando transport mapping, scalable catalyst synthesis, and programmable catalytic nanoarchitectures. This review offers a predictive design strategy for next-generation multifunctional catalytic materials by focusing on the integrated transport-reaction system instead of only focusing on the structure. Full article

Influence maximization (IM) in social networks aims to identify the optimal set of seed nodes that maximizes influence spread under a given diffusion model. The standard Grey Wolf Optimizer (GWO) suffers from two fundamental limitations when applied to this problem: an inflexible exploration–exploitation transition controlled by a linearly decreasing parameter; and a rigid three-level leadership hierarchy that suppresses individual diversity and promotes premature convergence. In this paper, we propose a Multi-Role Cooperative Grey Wolf Optimizer (Multiple-roles GWO) that addresses both limitations through two complementary mechanisms. First, a Q-learning-based adaptive phase transition mechanism monitors population diversity, fitness improvement rate, and iteration progress in real time, enabling the algorithm to dynamically shift between exploration and exploitation. Second, inspired by the principle of division of labor, the exploitation phase is restructured into a four-role cooperative framework comprising leaders, explorers, followers, and losers, each executing a distinct search strategy to improve local search coverage and maintain population diversity. Experiments on six real-world social networks under the Independent Cascade model show that Multiple-roles GWO achieves competitive or superior influence spread compared with state-of-the-art heuristic baselines, with comparable computational efficiency. Full article

Higher-order nonlinear effects induced by electrostatic coupling, large-deflection deformation and bending stiffness substantially alter the quenching evolution of elastic membranes in practical electrostatic MEMS devices. This paper investigates a higher-order electrostatic MEMS model with nonlinear diffusion power $m\in {\mathbb{N}}^{+}$. Stronger electrostatic nonlinearities modify quenching behavior, and the parity of m significantly influences the construction of discrete schemes. We first construct a semidiscrete scheme for odd-power cases, while the even-power cases require a fully discrete scheme with adaptive time-stepping near the quenching time. For both schemes, we establish stability and convergence of the corresponding discrete quenching solutions. Finally, numerical simulations corroborate the theoretical findings and provide intuitive illustrations of the quenching behavior, clearly demonstrating how higher-order nonlinearities affect the evolution of quenching, with insights for defining safe operating ranges and mitigating device failure. Full article

The sponge construction serves as a fundamental design framework for hash functions and authenticated encryption algorithms. In differential cryptanalysis of large-state permutations underlying such algorithms, conventional approaches are constrained by state size and diffusion speed. The feature extraction capability of deep learning offers a potential alternative to mitigate these limitations. To improve the distinguishing performance of differential neural distinguishers against sponge-based algorithms, a methodology integrating data construction and network architecture optimization is proposed. Specifically, a multi-sample triplet input format is designed to preserve differential characteristics, and a convolutional block attention module is introduced to capture long-range dependencies along both the channel and spatial dimensions within the large-state permutation. Experimental evaluations were conducted on the Keccak and Ascon algorithms. For Keccak, the maximum distinguishable round number was identified as 3. At this round number, Keccak-p achieved full distinguishability (100% accuracy), while the sponge-based SHA-3-256 attained a distinguishing accuracy of 99.99%, improving upon the previous best result by 0.95 percentage points. For Ascon, the maximum distinguishable round number was 4, where Ascon-p achieved an accuracy of 54.85% with 64 sample pairs—the highest reported accuracy for this setting—while delivering comparable performance at the matched 32-pair setting (53.40% vs. 53.54% in prior work) with approximately one-twelfth of the training epochs; under the same setting, the sponge-based Ascon-Hash256 achieved an accuracy of 53.06%. These findings demonstrate the effectiveness of the proposed framework in enhancing neural distinguisher accuracy against sponge-based algorithms and offer an analytical approach for empirical security evaluation, with results qualitatively consistent with the indifferentiability bound of the sponge construction. Full article