Search Results (2,714)

Cloud-based intrusion detection systems (IDSs) increasingly rely on deep learning classifiers to identify malicious traffic; however, this reliance exposes them to adversarial evasion attacks in which adversaries craft near-imperceptible perturbations to bypass detection. Existing defenses based on conventional adversarial training often recover robustness against known perturbation patterns at the cost of degraded detection accuracy on canonical attack categories—a robustness–accuracy trade-off that remains an open challenge in the field. In this paper, we propose GT-CSAT (Game-Theoretic Cost-Sensitive Adversarial Training), a novel defense framework tailored for cloud security environments. GT-CSAT couples an improved Wasserstein GAN with Gradient Penalty (WGAN-GP) threat generator—conditioned on attack semantics to simulate functionally consistent and highly covert traffic variants—with a minimax adversarial training loop governed by a game-theoretic cost-sensitive loss function. The proposed loss function assigns asymmetric misclassification penalties derived from a two-player zero-sum payoff matrix, enabling the detector to maintain vigilance over both novel adversarial variants and well-characterized conventional threats simultaneously. Specifically, misclassifying an adversarially perturbed attack as benign incurs a strictly higher penalty than the symmetric cross-entropy baseline, while the cost weights are dynamically adapted via a Nash equilibrium-inspired update rule during training. We conduct comprehensive experiments on the Cloud Vulnerabilities Dataset (CVD), CICIDS-2017, and UNSW-NB15, which encompass diverse cloud-specific attack scenarios including denial-of-service, port scanning, brute-force, and SQL injection traffic. Under six representative evasion strategies—FGSM, PGD, C&W, BIM, DeepFool, and IDSGAN-style black-box perturbations—GT-CSAT achieves an average robust accuracy of 94.3%, surpassing standard adversarial training by 6.8 percentage points and the undefended baseline by 21.4 percentage points, while preserving clean-traffic detection at 97.1%. These results confirm that the game-theoretic cost structure effectively decouples robustness from accuracy, yielding a Pareto-superior detection profile relative to competing baselines across all evaluated threat models. The source code and experimental configurations have been publicly released to facilitate reproducibility. Full article

Accurate classification of physical activity (PA) intensity is essential for exercise prescription, rehabilitation monitoring, and evaluation of guideline adherence. However, widely used wrist-worn accelerometer cut-points may substantially misclassify physiological intensity. This study evaluated absolute accelerometer thresholds during a maximal 2400 m run in military office workers and examined whether individualized cut-points improve agreement with physiological intensity. Seventy-four military office workers completed the test while wearing a wrist-worn ActiGraph GT9X Link and a chest-worn Zephyr BioHarness. Participants achieved near-maximal physiological effort, with peak heart rate averaging 187 ± 11 bpm (95 ± 4.2% age-predicted HRmax). Despite this high intensity, absolute wrist-worn cut-points classified only 34.5% of participants as performing vigorous activity for most of the test. Individualized cut-points, derived from each participant’s individual reference intensity, calculated as the three highest consecutive one-minute epochs during the 2400 m test, substantially improved agreement between accelerometer-derived classifications and physiological intensity. Agreement with %HRmax increased from fair (κ = 0.31), using absolute thresholds, to good (κ = 0.74), using individualized thresholds, and intraclass correlation increased from 0.52 to 0.81. These findings demonstrate that absolute cut-points markedly underestimate high-intensity activity, potentially leading to inaccurate exercise load monitoring and misinterpretation of training intensity. Individualized calibration during a standardized maximal running test provides a feasible strategy to improve the validity of intensity assessment using wearables. Although the study population consisted of military office workers, the approach may be applicable to other active populations. However, further validation in independent samples is needed. Full article

For over a decade, non-integrating Sendai virus vectors have been the gold standard for induced pluripotent stem cell (iPSC) reprogramming. However, as the field shifts toward regenerative and precision medicine and large-scale biorepositories, Sendai virus workflow necessitates dedicated viral-clearance testing, specialized manufacturing controls, and heightened regulatory oversight, leading to increased cost. While mRNA-based reprogramming offers a non-viral alternative, traditional mRNA delivery methods like electroporation are often physiologically disruptive. This study evaluates an mRNA-reprogramming platform that delivers lipid nanoparticles (mRNA-LNPs) via receptor-mediated endocytosis. By utilizing both Sendai virus and mRNA-LNP approaches to reprogram PBMCs from the same donor, we established a genetically identical starting point. Results demonstrate that mRNA-LNP-reprogrammed iPSCs maintain genomic integrity, retain the donor KCNH2 c.2398+5G>T variant, and exhibit characteristic colony morphology, pluripotency markers, and trilineage differentiation capacity consistent with the Sendai-reprogrammed counterparts. The mRNA-LNP-reprogrammed iPSCs differentiate into iPSC-derived cardiomyocytes presenting sarcomeric structures and electrophysiological activity, recapitulating a disease-specific phenotype. Notably, the mRNA-LNP workflow reached these milestones in significantly fewer passages than the Sendai virus workflow, markedly shortening timelines and reducing costs. These findings highlight mRNA-LNP reprogramming as a potentially attractive and effective, virus-independent platform to support future regenerative and precision medicine initiatives and scalable biobanking. Full article

This study aimed to evaluate the compensatory effects of amino acid (AA) supplementation on laying hens fed a low-protein diet, focusing on production performance, egg quality, blood immunity, total tract retention, and intestinal health. A total of 180 Hy-Line Brown laying hens (35 weeks old) were randomly allocated to three dietary treatments with 12 replicates (5 hens/replicate): (1) control diet (15.90% crude protein, CON), (2) reduced-protein diet (15.20% crude protein, NG), and (3) reduced-protein diet supplemented with 400 g/t methionine, 400 g/t lysine, and 300 g/t threonine (15.20% crude protein, LAA). Over the 12-week experimental period, production parameters were monitored weekly. The results showed that compared to the NG group, AA supplementation significantly increased laying rates during weeks 1–4 (p < 0.05) and reduced the feed-to-egg ratio during weeks 9–12 (p < 0.05), with a 3.10% decrease in feed conversion ratio (FCR). By week 12, the NG group exhibited lower Haugh unit values than both the CON and LAA groups (p < 0.05), indicating improved egg quality with AA supplementation. Serum analysis revealed elevated immunoglobulin Y levels and upregulated expression of the anti-inflammatory cytokine IL-10 in the LAA group (p < 0.05). Total tract retention of crude protein (CP) and phosphorus (P) was significantly enhanced in the LAA group compared to CON and NG groups (p < 0.05). However, in terms of cecal microbiota, no significant differences were observed among the experimental groups. These findings demonstrate that AA supplementation alleviates the adverse effects of low-protein diets by enhancing nutrient utilization and immune modulation, thereby improving productive performance, egg quality, and total tract retention in laying hens. Full article

²⁰Na is a well-known $\beta$-delayed $\alpha$ emitter, owing to the large decay energy of ²⁰Na above the α + ¹⁶O threshold in the $A=5\alpha$ daughter nucleus ²⁰Ne. In this work, the decay property of ²⁰Na is investigated in detail via the β-γ β-α and β-γ-α coincidence spectroscopy. As the day-one experiment of the Beijing Rare Isotope Facility (BRIF), the intense ²⁰Na beam was produced using the Isotope Separator On Line (ISOL) technique through the 100 MeV proton bombarding a stack of MgO as a thick target. Specific interest was focused on the exotic decay mode of ²⁰Na; the previously reported low-energy α lines at 713 and 846 keV were confirmed, and several weak β-γ-α decay sequences were clearly identified for the first time, thanks to the strong resolving power of α-γ coincidence spectroscopy. The decay properties of ²⁰Na are compared to the shell model calculation, which agree reasonably well with the allowed β transition strengths and subsequent electro-magnetic transitions with the use of the sd shell-model space with the USDB interaction. Full article

Background and Objectives: Thrombocytopenia complicates 6.6–11.6% of pregnancies. While gestational thrombocytopenia (GT) is usually benign, etiologies such as immune thrombocytopenia (ITP), preeclampsia, and HELLP syndrome require individualized management. This study aimed to characterize the etiological spectrum, maternal peripartum hematologic outcomes, blood product utilization, and mode of delivery in a tertiary-center cohort of thrombocytopenic pregnancies and to assess whether platelet count should influence delivery mode decisions. Materials and Methods: This retrospective cohort study included 137 thrombocytopenic pregnant women at a tertiary center (2010–2019), categorized by etiology and severity. Peripartum hemoglobin, hematocrit, and platelet counts were compared between delivery groups. Blood product utilization was recorded and analyzed using t-test, ANOVA, chi-square, Fisher’s exact, and Fisher–Freeman–Halton tests; binary logistic regression was used for multivariable analysis. Results: GT (43.1%) and ITP (32.1%) were the most prevalent diagnoses; cesarean delivery rate was 52.6%. Postpartum Hb was higher in the vaginal delivery group (10.24 ± 1.28 vs. 9.80 ± 1.26 g/dL; p = 0.003), while platelet counts were paradoxically lower (p = 0.039). Platelet transfusion rates did not differ significantly between delivery modes (23.1% vs. 27.8%; p = 0.621). Severe thrombocytopenia required platelet transfusion in 92.6% of cases versus 11.6% (moderate) and 0% (mild) (p < 0.001). RBC transfusion was highest in gestational hypertensive disease (41.2%) versus GT (5.1%) and ITP (2.3%) (p < 0.001). General anesthesia was used in 75% of cesarean cases. Conclusions: Delivery mode in thrombocytopenic pregnancies should be guided by obstetric indications, not platelet count alone. Although postpartum platelet counts declined more steeply after vaginal delivery, this did not increase transfusion requirements. Gestational hypertensive disorders carried the greatest hemorrhagic burden, highlighting the need for etiology-specific multidisciplinary planning. The high general anesthesia rate warrants prospective institutional audit of anesthetic decision-making protocols to determine adherence to current neuraxial anesthesia thresholds. This study is limited to maternal peripartum hematologic outcomes; neonatal outcomes were not captured and should be addressed in future prospective research. Full article

Fluorapatite is a typical phosphate rock resource. Fluorapatite tends to generate fine mud agglomeration, which induces dehydration challenges owing to its inherently fine particle size and negative surface charge. In this paper, phosphate tailings slurries from a phosphate mine in Hubei Province, China, were selected as the research object, and flocculation–dehydration experiments were conducted using anionic, cationic, and nonionic polyacrylamide (PAM) flocculants. The results show that the maximum settling velocity is 51 mm/s and the moisture content of filter cake is 41.54%, which were obtained when the unit consumption of cationic flocculant with molecular weight 12 million was 1000 g/t. The mechanism of sedimentation and dehydration was studied by infrared spectroscopy and a particle size analyzer. The results showed that polyacrylamide was effectively adsorbed on the mineral surface, and the size of flocs increased significantly. Finally, the mechanism of sedimentation and dehydration was proposed. It has important guiding significance for the efficient solid–liquid separation and water circulation of fluorapatite mineral processing wastewater. Full article

The rapid advancement of high-technology industries critically depends on the supply of rare earth elements (REEs), which constitute strategic raw materials for knowledge-intensive sectors. This study proposes an integrated methodological framework for the development of the Kundybay REE deposit, combining geomechanical substantiation of open-pit slope stability with beneficiation optimization of weathering-crust ores. The novelty of the research lies in the justification of rational pit-wall parameters based on a coupled 3D geological–geotechnical model implemented in the Micromine environment. Structurally disturbed zones were identified and subjected to detailed stability analysis using numerical modeling techniques. Particular emphasis was placed on the delineation of geotechnical domains and factor-of-safety evaluation in Slide2, which enabled refinement of design solutions and optimization of slope geometry. Simultaneously, the study addresses processing challenges of refractory weathering-crust ores through the application of ultrasonic aerohydraulic desliming combined with a vibro-centrifugal separation device. Optimization of slope parameters resulted in a 30 Mt reduction in stripping volume, corresponding to a 16.7% economic gain. The proposed flowsheet achieved a concentrate grade of 948.89 g/t REEs with an overall recovery of 45.66%. The results confirm that integrating geomechanical design with process engineering provides a technically robust and economically efficient basis for REE deposit development. Full article

Phosphogypsum (PG), a major by-product of the phosphate industry, has potential for improving acidic and nutrient-poor red soils, yet its agronomic benefits and heavy metal risks require systematic evaluation. A field experiment was conducted with five treatments, CK (soil only), GT (50% modified phosphogypsum, MPG), TT (40% MPG), ZT (50% phosphorite tailings), and DT (25% MPG + 25% lake sediment), to assess their effects on soil properties, enzyme activities, peanut growth, yield, quality, and heavy metal accumulation. All amendments improved soil structure, moisture retention, nutrient availability, and enzymatic activities. Peanut pod and kernel yields increased under all treatments, with DT achieving the greatest improvements (29.89% and 40.88%, respectively), whereas ZT showed the weakest response (1.91% and 6.26%). DT also achieved the highest soil quality index, and performed best in both yield improvement and root development. Although Cd accumulation increased under DT, heavy metal concentrations in peanut kernels remained below national food safety limits. Overall, DT was identified as the most effective amendment for enhancing red soil fertility and peanut productivity, while long-term monitoring of Cd bioavailability is recommended to ensure sustainable and safe application. Full article

Gamma-tocotrienol (GT3) is one of the constituents of vitamin E that demonstrated significant radioprotective efficacy in murine and nonhuman primate (NHP) models. Considering the antioxidant activity of GT3 and its role in terminating lipid peroxidation, we hypothesize that mechanism of radioprotective effect of GT3 may involve the inhibition of irradiation-induced ferroptosis—a form of regulated cell death characterized by excessive, iron-dependent, peroxidation of lipids in cellular membranes. To test this hypothesis, the metabolomic and proteomic data from serum samples of GT3- or vehicle-treated NHPs exposed to 12 Gy (partial- or total-body) radiation was analyzed with focus on lipid peroxidation markers and proteins involved in iron metabolism. Four secondary lipid peroxidation products were identified including 4-oxo-2-nonenal (4-ONE), 4-hydroperoxy-2-nonenal (4-HPNE), 3,4-epoxynonanal (3,4-ENA), and trans-4,5-epoxy-(2E)-decenal (4,5-EDE). In vehicle-treated animals, their concentrations increased significantly as soon as 4 h after irradiation and then gradually declined. GT3 treatment mitigated this radiation-induced increase. In addition to lipid peroxidation products, similar patterns of change were observed for several polyunsaturated, monounsaturated, and saturated fatty acids as well as amino acids such as lysine and its derivatives. Taken together, these metabolomic changes suggest that irradiation induces cellular membrane damage through enhanced lipid peroxidation, while GT3 exerts a protective effect against this process. In addition, GT3 increased serum levels of haptoglobin and hemopexin—two plasma scavenger proteins that play complementary protective roles in iron and heme homeostasis. Although the present study does not conclusively demonstrate that GT3 mediates radioprotection via inhibition of ferroptosis, the data suggest that GT3 limits membrane damage and reduces susceptibility to ferroptosis by enhancing iron and heme scavenging. Further investigation into the interaction between GT3 and key components of ferroptosis following exposure to ionizing radiation is therefore warranted. Full article

Feature-based visual–inertial odometry (VIO) often suffers from initialization failures and tracking drift under degraded visual conditions, such as low-texture regions, abrupt illumination changes, and scenes with a high ratio of dynamic correspondences. We present RISE-VIO, a real-time inertial-navigation-system-centric (INS-centric) visual–inertial odometry system that improves robustness by introducing GNC-style robustification into two failure-critical stages: initialization and per-frame pose estimation. For robust initialization, we develop a GNC-based decoupled rotation–translation initialization module with a two-stage observability gate, consisting of (i) rotation-compensated parallax-rate screening and (ii) a spectral-stability test on the linear global translation (LiGT) system. For online robustness, we design an IMU-prior-guided GNC-EPnP module to selectively downweight or reject outlier correspondences during pose estimation. Experiments on public benchmark datasets show that RISE-VIO achieves more reliable initialization and more stable trajectory estimation in challenging visual conditions while maintaining real-time performance. Additional Monte Carlo perspective-n-point (PnP) evaluations further support the robustness of the proposed pose estimation module under severe outlier contamination. Full article

Many challenges are commonly encountered in the underground mining of steeply dipping thin-to-medium-thick orebodies associated with weak hanging wall rockmass, such as stope back collapse, high ore dilution, and poor stoping stability. To address these issues, a synergistic mining method combining sidewall retaining and open stoping with a delayed backfilling method is proposed. Taking the north wing orebody of the Erlihe lead–zinc mine as the engineering background, a 3D finite element numerical simulation model was established using MIDAS GTS(2026 version) to conduct a comparative analysis between the proposed mining method and the current mining method. The mechanical response characteristics of crown pillar stress, crown pillar settlement, hanging wall displacement, and plastic zone evolution were systematically investigated under different mining stages. The results show that the proposed method improves the stress and deformation distribution at the bottom of the crown pillar. The peak stress decreases from 13.72 MPa to 12.86 MPa, and the spatial extent of the high-stress zone is noticeably reduced. Meanwhile, the maximum crown pillar subsidence decreases, while the width of the main subsidence zone decreases from 11 nodes to 9 nodes, and the settlement of the end region decreases by 6.05%. In terms of hanging wall response, the maximum displacement is reduced by 9.3–26.5% during the stope extraction stage and 9.6–10.0% during the inter-pillar recovery stage, with an overall average reduction of approximately 14.0%. Furthermore, the plastic zone in the hanging wall surrounding rock becomes smaller and develops later under the proposed mining method. Our findings demonstrate that the new proposed mining method effectively modifies the stress transfer path, mitigates deformation of both the crown pillar and hanging wall rock, and delays the development of plastic failure, thereby improving stope stability under weak hanging wall rockmass conditions. The proposed method provides a practical technical solution for the safe and efficient extraction of steeply dipping thin-to-medium-thick orebodies. Full article

Background/Objectives: Peptide receptor radionuclide therapy (PRRT) is an established treatment for metastatic neuroendocrine tumors (NETs), yet long-term disease control occurs only in a subset of patients. Predicting progression-free survival (PFS) could support individualized treatment planning. This study evaluates laboratory, imaging, and multimodal deep learning models for PFS prediction in PRRT-treated patients. Methods: In this retrospective, single-center study 116 patients with metastatic NETs undergoing [177Lu]Lu-DOTATOC were included. Clinical characteristics, laboratory values, and pretherapeutic somatostatin receptor positron emission tomography/computed tomographies (SR-PET/CTs) were collected. Seven models were trained to classify low- vs. high-PFS groups, including unimodal (laboratory, SR-PET, or CT) and multimodal fusion approaches. Performance was assessed via repeated 3-fold cross-validation with area under the receiver operating characteristic curve (AUROC) and area under the precision–recall curve (AUPRC). Explainability was evaluated by feature importance analysis and gradient based saliency maps. Results: Forty-two patients (36%) displayed short PFS (≤1 year) and 74 patients displayed long PFS (>1 year). Groups were similar in most characteristics, except for higher baseline chromogranin A (p = 0.003), elevated $\gamma$-GT (p = 0.002), and fewer PRRT cycles (p < 0.001) in short-PFS patients. The Random Forest model trained only on laboratory biomarkers reached an AUROC of 0.59 ± 0.02. Unimodal three-dimensional convolutional neural networks using SR-PET or CT performed worse (AUROC 0.42 ± 0.03 and 0.54 ± 0.01, respectively). A multimodal fusion model integrating laboratory values, SR-PET, and CT—augmented with a pretrained CT branch—achieved the best results (AUROC 0.72 ± 0.01, AUPRC 0.80 ± 0.01). Explainability analyses provided insights into model predictions, with explainability patterns in the fusion model appearing physiologically plausible and predominantly tumor-focused. Conclusions: Multimodal deep learning combining SR-PET, CT, and laboratory biomarkers outperformed unimodal approaches for PFS prediction after PRRT. Upon external validation, such models may support risk-adapted follow-up strategies. Full article

Pinoresinol diglucoside (PDG), one of the major lignans isolated from E. ulmoides Oliver bark, has various pharmacological functions, including antihypertension and prevention of osteoporosis. However, the glycosyltransferase-encoding gene (GT) involved in regulating the glycosylation of pinoresinol to form PDG has not been reported in E. ulmoides. In this study, we screened and cloned the EuGT8 gene from E. ulmoides based on our transcriptome data. The expression pattern of the EuGT8 gene exhibited a strong positive correlation with dynamic changes in the PDG contents in three different organs of E. ulmoides. The expression level of the EuGT8 gene and PDG content were significantly decreased in asODN-EuGT8-treated shoot tips in comparison with the control group. Prokaryotic expression of the EuGT8 gene revealed that the purified EuGT8 protein could catalyze the conversion of pinoresinol into PDG. In addition, we performed transcriptional and metabolomic analyses to compare the differences between transgenic Arabidopsis and WT plants. A total of 1799 DEGs and 294 DEMs were identified in transgenic and WT plants. KEGG enrichment analysis showed that the DEGs were mainly enriched in phenylpropanoid biosynthesis, secondary metabolite biosynthesis, and starch/sucrose metabolism pathways. The DEMs were mainly enriched in ABC transporters, aminoacyl-tRNA biosynthesis, biosynthesis of amino acids, phenylpropanoid biosynthesis, and flavone and flavonol biosynthesis pathways. Correlation analysis between DEGs and DEMs identified a total of 231 DEGs associated with 38 DEMs, which were mainly distributed in multiple metabolic pathways. This finding provides both theoretical insights and genetic resources for breeding high-PDG E. ulmoides varieties, facilitating marker-assisted selection (MAS) and promoting sustainable E. ulmoides production in Guizhou. Full article

Accurate trajectory tracking of redundant manipulators is difficult because the controller must simultaneously model local couplings between adjacent joints and global dependencies across the whole kinematic chain. Existing reinforcement learning methods typically employ multilayer perceptrons, which do not explicitly exploit manipulator structure and therefore show limited stability and representation ability in high-dimensional continuous control tasks. This paper proposes GT-TD3, a Graph Transformer-enhanced-Twin Delayed Deep Deterministic Policy Gradient framework, for redundant manipulator trajectory tracking. The proposed actor first converts the raw system state into joint-level node features and uses a graph neural network to extract local kinematic coupling information. A Transformer is then employed to capture long-range dependencies among joints. To strengthen the use of structural priors, topology- and distance-related bias terms are incorporated into the attention mechanism, enabling the network to encode manipulator structure during global feature learning. Experiments on a 7-DoF KUKA iiwa manipulator in PyBullet demonstrate that GT-TD3 outperforms MLP, pure GNN, and pure Transformer baselines in tracking performance. The proposed method achieves more stable training, faster convergence, and smoother and more accurate end-effector motion. The results show that the integration of local graph modeling and structure-aware global attention provides an effective solution for high-precision trajectory tracking of redundant manipulators. Full article