搜索结果：Cisco

Historical data can provide critical ecological information for species across the globe, many of which are facing unprecedented rates of ecosystem change. Yet, historical information related to freshwater species, especially fishes, remains scattered, often in original formats, and underutilized for informing conservation and restoration activities. Here, we present a Data Descriptor called Coregonine Spawning History (CORHIST), a database designed to house diverse data related to past spawning and nursery areas for fishes in the family Salmonidae, subfamily Coregoninae (ciscoes and whitefishes), in the Laurentian Great Lakes and their tributaries. Data for 11 species of coregonines historically occurring in the Great Lakes are included in CORHIST. Over 3,400 occurrence records at the coordinate scale have been entered, over 2,200 of which are for Cisco (Coregonus artedi) and Lake Whitefish (C. clupeaformis)-two focal species for which there is either multinational conservation interest or restoration efforts underway in the Laurentian Great Lakes. CORHIST is already proving useful for several studies developing habitat suitability models and delineating spatial units for conservation or restoration planning.

There are few data on the longer-term illness trajectory of patients following hospitalisation for COVID-19. We prospectively enrolled 267 adults hospitalised for COVID-19. Longer-term follow up was available for 260 participants. Event rates for death or unplanned hospitalisation were calculated using a Poisson model. Univariate and multivariable analyses identified baseline predictors, with a backward selection process for the best fitting model. The mean age of COVID-19 participants was 54.9±12.1 years, and 41% were female. During median follow-up of 1028 days (IQR:1000,1085), 112 individuals (43.1%) had at least one event including 6 deaths (2.3%). There were 252 events in total. The first event rate was 18.9 per 100 person-years (95%CI: 15.7, 22.8). Multivariable predictors included healthcare worker status (HR 0.59, 95%CI: 0.34, 1.02, p=0.046), Charlson Comorbidity Index (HR 1.13, 95%CI: 1.02, 1.24, p=0.020), current smoking (HR 2.49, 95%CI: 1.21, 5.11, p=0.010), and haemoglobin (HR 0.93, 95%CI: 0.88, 0.99, p=0.020). The WHO Clinical Severity Score was not a significant predictor (p=0.187). Comorbidity, current smoking status and haemoglobin predict illness trajectory following hospitalisation for COVID-19, rather than illness severity during hospitalisation. Further research is needed to explore interventions targeting these factors to improve prognosis. CISCO-19; http://NCT04403607. Registration date; 23/05/2020 The online version contains supplementary material available at 10.1186/s12879-025-12487-w.

Raised cardiac troponin-I is a common finding in patients hospitalised with acute viral infections, including but not limited to COVID-19. This often occurs in the absence of overt myocardial injury presenting a challenge for interpretation. The mechanisms underlying troponin elevation are uncertain. The CISCO-19 (Cardiovascular Imaging in SARS-CoV-19) study (NCT04403607) is a prospective, multicentre cohort study, in which hospitalised PCR-confirmed COVID-19 participants (N=267) underwent multisystem evaluation at enrolment and at 28-60&#x2009;days. The study incorporated plasma proteomics (SOMAscan V.4.1), cardiovascular MRI and clinical biomarkers. Of these, 211 had baseline plasma proteomic data and 185 completed follow-up sampling. Matched proteomic and imaging data were available for 155 participants (mean age: 55 years (SD 12); 43% female). A high likelihood of myocarditis was identified in 13.2% (N=21/159) of participants. High-sensitivity troponin-I was modestly elevated at enrolment (median 3&#x2009;ng/L; IQR 2-6; n=159). Among males (n=90), 9.3% had a high-sensitivity troponin that exceeded 34&#x2009;ng/L. Among females (n=69), 4.5% exceeded 16&#x2009;ng/L. Smooth muscle myosin light chain proteins were downregulated at follow-up (log2 fold change -0.12 to -0.6; all adjusted p<0.02) and positively correlated with high-sensitivity troponin-I, but not N-terminal brain natriuretic peptide or cardiac MRI indices (n=155). Troponin elevation, exemplified here by COVID-19, could reflect systemic vascular injury. Recognising this mechanism may refine interpretation of cardiac biomarkers in viral illness and supports the investigation of vascular injury in future therapeutic strategies and biomedical studies.

Walnut (Juglans regia L.) exhibits a high sensitivity to water deficit, making it crucial to comprehend this characteristic in order to optimize irrigation strategies to improve its productivity. Deficit irrigation is widely used under drought conditions to achieve water savings goals. This study examines the impact of sustained deficit irrigation (SDI) strategies-applying 33%, 50%, or 75% of the crop water demand-on yield and quality parameters of two walnut cultivars (Chandler and Cisco) over a three-year monitoring period. These treatments were compared against control trees receiving full irrigation at 100% of crop water requirements (C100). The nut yield was significantly and proportionally reduced under the SDI treatments. In the experiment, the average yield for cv. Chandler amounted to 6.7, 6.4, and 12.2 kg tree-1 under SDI33, SDI50, and SDI75, respectively, which were less than 13.9 kg tree-1 in the C100 plot. Similarly, cv. Cisco yielded 8.0, 11.6, 11, and 15.6 kg tree-1 under SDI33, SDI50, SDI75, and C100, respectively. These findings indicate that the cultivar Cisco exhibits greater tolerance to moderate and intermediate levels of water deficit. Furthermore, the SDI treatments notably influenced several morphological and physicochemical kernel parameters. The key affected attributes include the weight, size, color, profiles of specific sugars, and mineral content (notably potassium, iron, and zinc), as well as the composition of unsaturated fatty acids (palmitoleic and cis-vaccenic) and polyunsaturated fatty acids (linoleic and &#x3b1;-linolenic), with pronounced effects observed particularly under the SDI75 treatment. Thus, deficit irrigation did not drastically affect the kernel quality parameters, and it is also possible to augment them by selecting the appropriate water stress level. Therefore, for both walnut cultivars, approximately 25% of the irrigation water (SDI75), equivalent to an average of 1681 m3 ha-1, can be conserved relative to the total crop water requirement without negatively impacting walnut tree performance in the short to medium term. Here, we show the key role of adjusting irrigation practices by stressing the benefits of SDI that can save water, foster water productivity, and boost walnut health-promoting phytochemicals.

The rapid expansion of consumer electronics has created an urgent need for advanced solutions to critical challenges such as predictive maintenance, user personalization, and device security. Traditional models often struggle to address these issues due to their limited adaptability and performance. This paper introduces GenAI-A, an innovative AI model that integrates Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), Dynamic Recommendation Algorithms (DRA), and anomaly detection techniques to provide a comprehensive solution for consumer electronics. By leveraging the strengths of generative AI and self-renewal capabilities, GenAI-A enhances predictive maintenance, optimizes user experience, and strengthens biometric security. The model&#x2019;s hybrid architecture utilizes GANs for realistic data representation, VAEs for the generation of complex data distributions, and DRA for real-time user personalization. The novelty of GenAI-A lies in its cross-regularized coupling between the GAN and VAE modules, where latent features are jointly optimized through a shared loss function to achieve consistent generative&#x2013;representational learning. Unlike conventional hybrids that treat these models independently, GenAI-A introduces a dynamic feedback mechanism in which the DRA and anomaly detection modules operate directly in the shared latent space, enabling self-adaptive personalization and continual refinement of generative outputs. Experimental validation across four real-world datasets, i.e., Smartphone Sensor, Labelled Faces in the Wild (LFW), Pecan Street Energy Consumption, and SECOM Manufacturing, demonstrates significant improvements in device uptime, user engagement, and biometric security, with a notable reduction in false positives. Unlike existing hybrid generative models, GenAI-A introduces a novel integration of these components in a dynamic, self-learning system that adapts in real time to evolving user behaviors and device conditions. This unique combination of techniques sets GenAI-A apart from traditional approaches, establishing a new benchmark for AI-driven solutions in consumer electronics.

Oligodendrocyte maturation and myelination are critical processes in human neurodevelopment, and their dysregulation is linked to numerous neurological disorders. While model organisms have provided insight into these processes, human-specific regulatory mechanisms remain poorly understood. This study investigated human THAP9, a protein homologous to the Drosophila P-element transposase, whose function in oligodendrocytes remains unknown. An analysis of RNA-sequencing data and H3K27ac ChIP-sequencing data from oligodendrocyte progenitor cells (OPCs) and mature oligodendrocytes (MOs) revealed significant upregulation of THAP9 during oligodendrocyte maturation. Co-expression analysis demonstrated a strong correlation with established markers of oligodendrocyte development, including myelin-associated genes (MOG, MBP) and key transcriptional regulators (PDGFRA, SOX5, SOX6, SOX11). THAP9 lacks homologues in mice, highlighting potential human-specific mechanisms in oligodendrocyte development and emphasising the importance of studying species-specific factors in neurodevelopment. Our findings suggest that THAP9 is a novel human-specific regulator of oligodendrocyte maturation and opens new avenues for studying myelination disorders.

Walnut rootstocks are commonly used in California orchards to provide resistance to soil-borne pests and diseases. However, little information exists about the impact of commercial rootstock on the common scion's physiological response under drought. This is becoming increasingly important since walnuts are commonly cultivated in semi-arid regions where frequent and severe droughts require efficient water use. We previously reported that own-rooted walnut rootstocks (RX1, VX211 and Vlach) differ in their physiological performance under drought. Here, we evaluated whether similar water relations and performance are conferred to a common English walnut scion (Juglans regia cv. Cisco). To do so, we used a mini-lysimeter platform to continuously track soil moisture and transpirational water loss from trees. Along with the canopy's estimated leaf area, changes in canopy shape and texture were evaluated using deep learning as an independent method to analyze canopy response to water stress. In support of our recent findings, the scion grafted onto rootstock RX1 exhibited subtle improvements in physiological performance associated with higher transpiration and canopy conductance under well-watered condition compared to Vlach and VX211 rootstocks. Canopy conductance, texture, and shape were not significantly affected by rootstock under water stress. However, Cisco grafted onto RX1 exhibited higher leaf turgor and water use efficiency, and lower osmotic potentials under water stress. Our results suggest some subtle differences in water relations between the rootstock genotypes, and propose an efficient deep-learning method to screen canopies for water stress-induced response through image processing.

[Image: see text]

The opaque nature of black-box deep learning (DL) models poses significant challenges for intrusion detection systems (IDSs) in Internet of Things (IoT) networks, where transparency, trust, and operational reliability are critical. Although explainable artificial intelligence (XAI) has been increasingly adopted to enhance interpretability, its impact on detection performance and computational efficiency in resource-constrained IoT environments remains insufficiently understood. This systematic review investigates the performance of an explainable deep learning-based IDS for IoT networks by analyzing trade-offs among detection accuracy, computational overhead, and explanation quality. Following the PRISMA methodology, 129 peer-reviewed studies published between 2018 and 2025 are systematically analyzed to address key research questions related to XAI technique trade-offs, deep learning architecture performance, post-deployment XAI evaluation practices, and deployment bottlenecks. The findings reveal a pronounced imbalance in existing approaches, where high detection accuracy is often achieved at the expense of computational efficiency and rigorous explainability evaluation, limiting practical deployment on IoT edge devices. To address these gaps, this review proposes two conceptual contributions: (i) an XAI evaluation framework that standardizes post-deployment evaluation categories for explainability, and (ii) the Unified Explainable IDS Evaluation Framework (UXIEF), which models the fundamental trilemma between detection performance, resource efficiency, and explanation quality in IoT IDSs. By systematically highlighting performance-efficiency gaps, methodological shortcomings, and practical deployment challenges, this review provides a structured foundation and actionable insights for the development of trustworthy, efficient, and deployable explainable IDS solutions in IoT ecosystems.

The etiology of Alzheimer's disease (AD) remains under active debate. In this perspective, we explore the hypothesis that a primarily infection-caused chronic dysregulation and weakening of human innate immunity via the underexpression, degradation, and inactivation of innate immune proteins necessary for direct antimicrobial effects and regulation of host defense and autophagy could lead to AD. Key evidence relates to the fact that important innate immune proteins such as LL-37-which can bind A&#x3b2; and block amyloid formation-as well as Apolipoprotein E, antiviral interferons, and TNF-&#x3b1;&#xa0;can be degraded and deactivated by enzymes produced by the common oral anaerobic pathogen Porphyromonas gingivalis (Pg). Pg produces numerous virulence factors; of particular importance for AD are Pg's gingipain cysteine proteases. Deleterious effects of chronic Pg infection and gingipains include a systemic downregulation and paralysis of the interferon response, particularly the antiviral interferon-lambda response, which enables replication of endemic herpesviruses. The result is a chronic, low-level viral infectious assault on gut, nerves, and brain causing the production of A&#x3b2; antimicrobial peptides, accumulation of A&#x3b2; plaques, phosphorylation of Tau, progressive neuroinflammation, and neurodegeneration. The resultant innate immune system dysregulation, as an AD etiology, ties together the well-known amyloid cascade hypothesis and the infectious theory of AD into a unified explanation of the pathology and cause of AD. If this theory holds true, it suggests preventative approaches: (1) test for and eradicate Pg from oral flora, and/or directly deactivate the gingipains; and (2) reduce Herpesvirus exacerbations by the use of antiviral drugs and/or vaccines (e.g., Bacillus Calmette-Gu&#xe9;rin).

The present study was designed to examine the mental health challenges facing Ukrainian migrants exposed to the 2022 Russian invasion and who now reside in the United States, focusing on trauma, posttraumatic stress disorder, and emotional distress. Our study explored how anticipatory grief and perseverative worry intersect and contributed to migrants' emotional distress, which is further compounded by their ongoing emotional connection to Ukraine, where many relatives remain in danger. Using a general inductive approach, we analyzed data from interviews with eight Ukrainian war migrants in the U.S. interviews were conducted in Ukrainian or Russian and analyzed using Dedoose Version 9.2.005. The majority (five of eight) of participants discussed posttraumatic stress disorder symptoms and extreme emotional distress (six of eight) as a direct result of their exposure to war and violence, either firsthand or through loved ones. Participant statements suggested that emotional distress and strong emotional connection to Ukraine often co-occurred, with (a) 62.5% of individuals reporting both concerns for the safety of family members still in Ukraine, (b) material/emotional connections to Ukraine, and (c) 75% of participants indicating that their emotional distress was highly linked to trauma and posttraumatic stress disorder. Ukrainian migrants face a complex spectrum of trauma and emotional distress, influenced by their ongoing connection to a conflict zone and by constant worry for their families' safety. (PsycInfo Database Record (c) 2026 APA, all rights reserved).

Head and neck adenoid cystic carcinoma (HNACC) is an uncommon but aggressive salivary gland malignancy with high recurrence and metastasis rates. Its "immune-cold" phenotype limits the efficacy of systemic therapy, and thus, radiotherapy is the principal treatment for unresectable cases. However, frequent radioresistance suggests the involvement of distinct molecular drivers. This study investigated the role of CCL5 in HNACC radioresistance, its mechanistic link to autophagy, and its potential as a radiosensitization target. Transcriptome sequencing of radiation-sensitive (SACC-83) and radiation-tolerant (SACC-LM) cell lines, integrated with GEO datasets, identified candidate genes. For functional validation, genetic manipulation and pharmacological inhibition with the CCL5/CCR5 antagonist maraviroc were performed. Cellular proliferation, invasion, apoptosis, cell cycle, and radiosensitivity were assessed. Autophagy was examined by transmission electron microscopy and western blotting, with a focus on AMPK/mTOR signaling. A xenograft mouse model was used to evaluate the therapeutic effect of maraviroc combined with radiotherapy. CCL5 expression was elevated in radioresistant cells and further induced by irradiation. High CCL5 levels promoted proliferation, migration, and invasion while reducing radiation-induced apoptosis and G&#x2082;/S arrest. Mechanistically, CCL5 activated AMPK, suppressed mTOR, and enhanced autophagy, supporting cell survival under radiation stress. CCL5 inhibition reduced autophagy, restored radiosensitivity, and synergized with irradiation, whereas autophagy activation reversed these effects. In vivo, maraviroc combined with radiotherapy decreased tumor burden, suppressed CCL5 expression, and inhibited autophagy. CCL5 promotes radioresistance in HNACC through maintenance of AMPK/mTOR-dependent autophagy. Targeting the CCL5/CCR5 axis enhances radiosensitivity and represents a promising therapeutic strategy. Not applicable.

Graph Neural Networks (GNNs) have emerged as a novel paradigm that enables scientists to model complex relational data in medical applications, offering unique advantages over traditional deep learning (DL) approaches for non-Euclidean domains. This paper provides a comprehensive review of current GNN architectures and their healthcare applications, with a focus on functional connectivity analysis, electrical-based diagnostics, and anatomical structure modeling. We analyze the strengths and limitations of spectral and spatial GNN variants, including Graph Convolutional Networks (GCNs), Graph Attention Networks (GATs), and spatio-temporal extensions. Based on our critical assessment of the state-of-the-art innovations, we propose several key directions for medical researchers actively developing GNN technology: (1) Dynamic graph representation learning to capture evolving physiological processes; (2) Multi-modal fusion techniques to integrate heterogeneous biomedical data streams; (3) Uncertainty-aware GNNs for robust clinical decision support; (4) Explainable GNN architectures to enhance interpretability for healthcare practitioners; and (5) Federated GNN frameworks to enable privacy-preserving collaborative learning across institutions. We also introduce a new Temporal Multi-modal Attention Graph Neural Network (TMA-GNN) architecture designed explicitly for longitudinal patient modeling and clinical trial optimization. Our TMA-GNN incorporates multi-head attention mechanisms, temporal edge construction, and a custom loss function to encourage temporal consistency in predictions. We introduce a conceptual framework for the Temporal Multi-modal Attention Graph Neural Network (TMA-GNN), which is designed to support disease progression modeling and clinical trial optimization in neurological disorders. Although the proposed model architecture is technically detailed, this manuscript focuses on the conceptual and methodological design, rather than presenting experimental results. By addressing these proposed research directions, we envision GNNs will play an increasingly pivotal role in precision medicine, disease progression modeling, and treatment personalization.

Background: Exposure to therapeutic landscapes has been consistently associated with reduced stress, improved affect, and enhanced emotion regulation among young adults. However, access to such environments is often limited on urban campuses where anxiety is prevalent. In response, this study conceptualizes the virtual therapeutic landscape (VTL) and proposes a design and evaluation model that translates therapeutic landscape theory into immersive virtual reality (VR). Methods: A three-stage mixed-methods design was employed. Semi-structured interviews (n&#x2009;=&#x2009;18) were thematically analyzed to identify core experiential dimensions that informed VTL model development. Expert analytic hierarchy process (AHP) ratings (n&#x2009;=&#x2009;12) yielded weights for domains and sub-criteria with acceptable consistency. On this basis, a standardized VTL exposure was administered to university students (n&#x2009;=&#x2009;60), who completed psychometric questionnaires. Text-mined features from open-ended responses were integrated with these outcomes to refine the expert-weighted model. Results: The calibrated model produced a coherent weighting structure, with sensory experience receiving the highest weight among the experiential dimensions. Brief VTL exposure significantly reduced state anxiety (p&#x2009;<&#x2009;0.001, d&#x2009;=&#x2009;1.040) and negative affect (p&#x2009;<&#x2009;0.001, d&#x2009;=&#x2009;0.570) and increased subjective vitality (p&#x2009;<&#x2009;0.001, d&#x2009;=&#x2009;0.794). Text mining supported an architecture in which sensory-narrative coupling and low-friction interaction experience act as primary levers, while personalization experience moderates these effects. Conclusions: This study develops a design and evaluation model for VTL targeting anxiety and emotion regulation in university students. Brief VTL exposure has shown measurable psychometric change; long-term effects and variation across VTL types remain priorities for future research.

Adversarial robustness in artificial intelligence is commonly defined in terms of input-level perturbations applied to static models. This study reconceptualises adversarial vulnerability for artificial and agentic AI systems by extending the threat model to autonomy, self-governance, and closed-loop decision-making, where behaviour unfolds dynamically through feedback and control. We develop a system-level analytical framework that formalises adversarial risk across perceptual, cognitive, and executive layers. The analysis is grounded in a PRISMA-compliant systematic literature review, bibliometric mapping, and targeted empirical validation. Established adversarial results from vision benchmarks and recent large-language-model red-teaming studies are synthesised to contextualise the framework, rather than to introduce new benchmark performance claims. The results demonstrate that no single defence mechanism provides robustness across all layers of agentic AI systems. Adversarial vulnerabilities propagate from perception to policy and actuation, with architectural similarity, domain shift, and feedback dynamics critically shaping transferability and failure modes. These effects have direct implications for safety-critical applications, including autonomous mobility, healthcare imaging, and biometric security. By framing higher-order agentic adversarial threats as hypothesis-driven, system-level risks, this work shifts adversarial AI security from benchmark-centric evaluation to behavioural integrity and lifecycle resilience. The proposed framework defines a coherent research agenda for agentic AI security that integrates control-theoretic reasoning and governance-aware defence design, addressing limitations of classical adversarial machine-learning theory.

暂无摘要（点击查看详情）

Agentic AI systems integrate foundation models, prompt templates, tool connectors, orchestration logic, and containerised dependencies, creating exploitability conditions that cannot be inferred from static Software Bills of Materials (SBOMs). Artificial Intelligence Bills of Materials (AIBOM) extend transparency to AI-specific artefacts, yet current CSAF/VEX workflows remain based on static component-CVE correlation without runtime validation. A protocol-driven framework is presented that binds SBOM and AIBOM artefacts to deterministic environment capture and structured runtime telemetry. Exploitability is computed from declared artefacts, observed activation conditions, and enforced execution policies. CSAF-VEX advisories are generated from combined static and runtime evidence, cryptographically signed, and validated through deterministic replay. Evaluation uses approximately 10,000 component entries across synthetic Agentic AI workloads (50-5,000 components), incorporating OSV, GitHub Advisory, KEV, and EPSS datasets. Under controlled experimental conditions, the framework achieves an F1-score of 0.93 (precision 0.96, recall 0.92), reduces false positives by up to 42% relative to static SBOM-CVE matching without runtime validation, and alters exploitability outcomes in 31% of AI-specific artefact cases through AIBOM extension. Advisory artefacts remain reproducible under deterministic replay. Binding AIBOM artefacts to runtime telemetry transforms CSAF-VEX generation from static disclosure into execution-grounded exploitability assessment for Agentic AI supply chains.

Federated learning (FL) is a distributed learning paradigm that facilitates training a global machine-learning model without collecting the raw data from distributed clients. Recent advances in FL have addressed several considerations that are likely to transpire in realistic settings, such as data distribution heterogeneity among clients. However, most of the existing works still consider clients' data distributions to be static or conforming to a simple dynamic, e.g., in participation rates of clients. In real FL applications, client data distributions change over time, and the dynamics, i.e., the evolving pattern, can be highly non-trivial. Furthermore, evolution may take place from training to testing. In this paper, we address dynamics in client data distributions and aim to train FL systems from time-evolving clients that can generalize to future target data. Specifically, we propose two algorithms, FedEvolve and FedEvp, which are able to capture the evolving patterns of the clients during training and are test-robust under evolving distribution shifts. FedEvolve explicitly models the temporal evolution by learning two distinct representation mappings that capture the transition between consecutive data domains for each client. In addition, FedEvp learns a single, evolving-domain-invariant representation by aligning current data with prototypes that are continuously updated from all previously seen domains. Through extensive experiments on both synthetic and real data, we show the proposed algorithms can significantly outperform the FL baselines across various network architectures.