Search Results (3,646)

This paper proposes a PI-Dual-STGCN fault diagnosis model based on a SHAP-LLM joint explanation framework to address issues such as the lack of transparency in the diagnostic process of deep learning models and the weak interpretability of diagnostic results. PI-Dual-STGCN enhances the interpretability of graph data by introducing physical constraints and constructs a dual-graph architecture based on physical topology graphs and signal similarity graphs. The experimental results show that the dual-graph complementary architecture enhances diagnostic accuracy to 99.22%. Second, a general-purpose SHAP-LLM explanation framework is designed: Explainable AI (XAI) technology is used to analyze the decision logic of the diagnostic model and generate visual explanations, establishing a hierarchical knowledge base that includes performance metrics, explanation reliability, and fault experience. Retrieval-Augmented Generation (RAG) technology is innovatively combined to integrate model performance and Shapley Additive Explanations (SHAP) reliability assessment through the main report prompt, while the sub-report prompt enables detailed fault analysis and repair decision generation. Finally, experiments demonstrate that this approach avoids the uncertainty of directly using large models for fault diagnosis: we delegate all fault diagnosis tasks and core explainability tasks to more mature deep learning algorithms and XAI technology and only leverage the powerful textual reasoning capabilities of large models to process pre-quantified, fact-based textual information (e.g., model performance metrics, SHAP explanation results). This method enhances diagnostic transparency through XAI-generated visual and quantitative explanations of model decision logic while reducing the risk of large model hallucinations by restricting large models to reasoning over grounded, fact-based textual content rather than direct fault diagnosis, providing verifiable intelligent decision support for industrial fault diagnosis. Full article

In achieving sustainable development goals, soils play a key role in environmental protection, natural resources, and food security. Peatlands are particularly important here, as they function at the interface between terrestrial and aquatic ecosystems and store large amounts of organic matter. However, organic soils are highly susceptible to transformation and degradation; therefore, their degradation caused by, among others, drainage properties is a high risk to both the environment and agriculture—it disrupts the ecosystems, causes greenhouse gas emissions, and eutrophicates the hydrosphere. Soil degradation in drained peatlands is associated with the transformation of soil organic matter (SOM), which in organic soils is the dominant component of the solid phase of the soil. The aim of our study was to assess the properties and degree of organic matter transformation in drained temperate peatland soils, with particular emphasis on sequential fractionation of SOM and humic acid properties. Due to the fact that in Poland, as many as 90% of non-forest peat bogs have been drained, we compare the mursh horizons that formed after peat bog drainage with the peat horizons that constitute the parent rock (where anaerobiosis occurs and morphological changes in the soil material are absent due to peat bog drainage). Studies were conducted on 11 soil profiles located in central-eastern Poland. Basic physicochemical soil properties were determined: pH, bulk density, contents of ash, SOM, total carbon (TC), and total nitrogen (TN). Sequential carbon fractionation was used to qualitatively analyze organic matter, which allowed for the identification of labile fractions, lipid fractions, humic substances (fulvic and humic acids), and residual fractions. Humic acids (HAs) were extracted using the Schnitzer method and analyzed for their elemental composition and spectrometric parameters in the VIS range. It was demonstrated that SOM transformation in drained temperate peatland soils was correlated with comprehensive changes in the soil’s physical and chemical properties. Compared to peat horizons, topsoil horizons were characterized by higher ash content and density, lower SOM content, and a lower TC/TN ratio. Qualitative SOM transformation during aerobic SOM transformation after draining the studied peatlands consisted of an increase in the amount of labile fractions and humic substances and a decrease in the lipid and residual fractions. The research results have shown that the HAs properties depended on the depth. HAs from topsoil horizons, compared to peat horizons, were characterized by a lower “degree of maturity,” as reflected by the values of atomic ratios (H/C, O/C) and absorbance coefficients (A_4/6 and ΔlogK). It was found that the share of the distinguished SOM fractions and HAs properties were closely correlated with the physical and chemical properties of the soils. The study demonstrated the usefulness of the sequential carbon fractionation method for assessing the effects of dewatered peat transformation. The obtained results could contribute to the development of good practices ensuring high quality of organic matter and stability of ecosystems, as well as to the development of methods for limiting the mineralization of organic matter (SOM), greenhouse gas emissions, and the loss of organic soils in agricultural areas. Full article

The increasing role of automation systems in energy-efficient buildings creates a need for simulation approaches that support standardized assessment already at the design stage. This paper presents a digital twin-based simulation framework that integrates building information modeling (BIM)-derived building data with MATLAB/Simulink models to enable regulation-oriented evaluation of building automation and control strategies. The proposed approach targets scenario-based analysis of automation maturity levels, covering conventional, advanced, and predictive configurations aligned with EN ISO 52120 and the Smart Readiness Indicator (SRI). A representative academic building model is used to demonstrate how the framework supports reproducible modeling of heating, ventilation, and air conditioning (HVAC), lighting, and shading control functions and enables consistent comparison of their energy-related behavior under unified boundary conditions. The results show that the framework effectively captures performance trends associated with increasing automation sophistication and reveals interaction effects between control subsystems that are not accessible in conventional energy simulation tools. The proposed methodology provides a practical and extensible foundation for early-stage, regulation-aligned evaluation of smart building solutions and for the further development of predictive and artificial intelligence (AI)-assisted control concepts. Full article

Oocytes cultured in vitro are exposed to high oxygen tension and lack follicular antioxidants, leading to redox imbalance. Eicosapentaenoic acid (EPA), a marine long-chain n-3 polyunsaturated fatty acid, possesses strong antioxidant activity. Here, using pigs as a model, we examined the effects of EPA on oocyte in vitro maturation (IVM) and subsequent developmental competence. Cumulus–oocyte complexes were cultured with EPA, followed by assessment of nuclear and cytoplasmic maturation and embryonic development; transcriptomic and proteomic analyses were conducted to explore underlying mechanisms. Supplementation with 10 µM EPA significantly improved maturation and blastocyst rates by reducing spindle defects, facilitating a more uniform organization of cortical granules and mitochondria. EPA increased resolvin E1 accumulation and reduced cumulus-cell apoptosis through downregulation of TNF-α and BAX and upregulation of BCL2. In MII oocytes, EPA lowered apoptosis, DNA damage, and ROS levels while enhancing SOD2 and GPX4 expression. Mitochondrial quality and turnover were improved via upregulation of PPARGC1A, NDUFS2, PINK1, LC3, FIS1, MUL1, and OPA1, alongside strengthened ER–mitochondria contacts. These findings demonstrate that EPA alleviates oxidative stress, optimizes mitochondrial function, and enhances porcine oocyte maturation and developmental competence in a parthenogenetic model, highlighting its potential as a marine-derived functional additive for reproductive biotechnology. Future studies will be required to validate these effects under fertilization-based embryo production systems and to further refine dose–response relationships using expanded embryo-quality endpoints. Full article

This investigation examines diffusion processes for predicting whole-stand volume, incorporating the variability and uncertainty inherent in regional, operational, and environmental factors. The distribution and spatial organization of trees within a specified forest region, alongside dynamic fluctuations and intricate uncertainties, are modeled by a set of nonsymmetric stochastic differential equations of a sigmoidal nature. The study introduces a three-dimensional system of stochastic differential equations (SDEs) with mixed-effect parameters, designed to quantify the dynamics of the three-dimensional distribution of tree-size components—namely diameter (diameter at breast height), potentially occupied area, and height—with respect to the age of a tree. This research significantly contributes by translating the analysis of tree size variables, specifically height, occupied area, and diameter, into stochastic processes. This transformation facilitates the representation of stand volume changes over time. Crucially, the estimation of model parameters is based exclusively on measurements of tree diameter, occupied area, and height, avoiding the need for direct tree volume assessments. The newly developed model has proven capable of accurately predicting, tracking, and elucidating the dynamics of stand volume yield and growth as trees mature. An empirical dataset composed of mixed-species, uneven-aged permanent experimental plots in Lithuania serves to substantiate the theoretical findings. According to the dataset under examination, the model-based estimates of stand volume per hectare in this region exhibited satisfactory goodness-of-fit statistics. Specifically, the root mean square error (and corresponding relative root mean square error) for the living trees of mixed, pine, spruce, and birch tree species were 68.814 m³ (20.4%), 20.778 m³ (7.8%), 32.776 m³ (37.3%), and 4.825 m³ (26.3%), respectively. The model is executed within Maple, a symbolic algebra system. Full article

At present, fruit picking mainly relies on manual operation. Taking the litchi (litchi chinensis Sonn.)-picking robot as an example, visual perception is often affected by illumination variations, low recognition accuracy, complex maturity judgment, and occlusion, which lead to inaccurate fruit localization. This study aims to establish an embodied perception mechanism based on “perception-reasoning-execution” to enhance the visual perception and decision-making capability of the robot in complex orchard environments. First, a Y-LitchiC instance segmentation method is proposed to achieve high-precision segmentation of litchi clusters. Second, a generative artificial intelligence model is introduced to intelligently assess fruit maturity and occlusion, providing auxiliary support for automatic picking. Based on the auxiliary judgments provided by the generative AI model, two types of dynamic harvesting decisions are formulated for subsequent operations. For unoccluded main fruit-bearing branches, a skeleton thinning algorithm is applied within the segmented region to extract the skeleton line, and the midpoint of the skeleton is used to perform the first type of localization and harvesting decision. In contrast, for main fruit-bearing branches occluded by leaves, threshold-based segmentation combined with maximum connected component extraction is employed to obtain the target region, followed by skeleton thinning, thereby completing the second type of dynamic picking decision. Experimental results show that the Y-LitchiC model improves the mean average precision (mAP) by 1.6% compared with the YOLOv11s-seg model, achieving higher accuracy in litchi cluster segmentation and recognition. The generative artificial intelligence model provides higher-level reasoning and decision-making capabilities for automatic picking. Overall, the proposed embodied perception mechanism and dynamic picking strategies effectively enhance the autonomous perception and decision-making of the picking robot in complex orchard environments, providing a reliable theoretical basis and technical support for accurate fruit localization and precision picking. Full article

Agricultural marketing increasingly integrates Agriculture 4.0 technologies—Blockchain, AI/ML, IoT, and recommendation systems—yet systematic evaluations of computational maturity and deployment readiness remain limited. This Systematic Literature Review (SLR) examined 99 peer-reviewed studies (2019–2025) from Scopus, Web of Science, and IEEE Xplore following PRISMA protocols to assess algorithmic performance, evaluation methods, and Technology Readiness Levels (TRLs) for agricultural marketing applications. Hybrid recommendation systems dominate current research (28.3%), achieving accuracies of 80–92%, while blockchain implementations (15.2%) show fast transaction times (<2 s) but limited real-world adoption. Machine learning models using Random Forest, Gradient Boosting, and CNNs reach 85–95% predictive accuracy, and IoT systems report >95% data transmission reliability. However, 77.8% of technologies remain at validation stages (TRL ≤ 5), and only 3% demonstrate operational deployment beyond one year. The findings reveal an “efficiency paradox”: strong technical performance (75–97/100) contrasts with weak economic validation (≤20% include cost–benefit analysis). Most studies overlook temporal, geographic, and economic generalization, prioritizing computational metrics over implementation viability. This review highlights the persistent validation–deployment gap in digital agriculture, urging a shift toward multi-tier evaluation frameworks that include contextual, adoption, and impact validation under real deployment conditions. Full article

Objective: We analyzed nanoparticle regulation research to examine the evolution of regulatory frameworks, identify major thematic structures, and evaluate current challenges in the governance of rapidly advancing nanotechnologies. By drawing parallels with the historical development of radiation regulation, the study aimed to contextualize emerging regulatory strategies and derive lessons for future governance. Methods: A total of 9095 PubMed-indexed articles published between January 2015 and October 2025 were analyzed using text mining, keyword frequency analysis, and topic modeling. Preprocessed titles and abstracts were transformed into a TF-IDF (Term Frequency–Inverse Document Frequency) document–term matrix, and NMF (Non-negative Matrix Factorization) was applied to extract semantically coherent topics. Candidate topic numbers (K = 1–12) were evaluated using UMass coherence scores and qualitative interpretability criteria to determine the optimal topic structure. Results: Six major research topics were identified, spanning energy and sensor applications, metal oxide toxicity, antibacterial silver nanoparticles, cancer nano-therapy, and nanoparticle-enabled drug and mRNA delivery. Publication output increased markedly after 2019 with interdisciplinary journals driving much of the growth. Regulatory considerations were increasingly embedded within experimental and biomedical research, particularly in safety assessment and environmental impact analyses. Conclusions: Nanoparticle regulation matured into a dynamic multidisciplinary field. Regulatory efforts should prioritize adaptive, data-informed, and internationally harmonized frameworks that support innovation while ensuring human and environmental safety. These findings provide a data-driven overview of how regulatory thinking was evolved alongside scientific development and highlight areas where future governance efforts were most urgently needed. Full article

As Thailand’s financial sector accelerates its digital transformation, cybersecurity has transitioned from a mere technical support function to a strategic imperative that governs operational risk and financial stability. This study empirically examines the efficacy of cyber risk controls and their correlation with perceived organizational readiness. Utilizing a quantitative survey of 53 specialized practitioners (N = 53), we assessed maturity across the six dimensions of the Bank of Thailand’s Cyber Resilience Assessment regulatory framework: Governance, Identification, Protection, Detection, Response, and Third-Party Risk Management. While descriptive statistics indicate high overall maturity ($\overline{\mathrm{x}}$ = 4.19, S.D. = 0.37), multiple regression analysis uncovers a critical “Protection Paradox”. Specifically, the “Protection” dimension exhibits a statistically significant negative impact on readiness (β = −0.432, p = 0.01), suggesting that over-engineered technical controls induce operational friction. In contrast, “Identification” emerged as the primary positive driver of readiness (β = 0.627, p < 0.01), highlighting visibility as a superior strategic lever. Furthermore, a structural disconnect was identified between strategic “Governance” and “Third-Party Risk Management” (r = 0.46), highlighting a “Silo Effect” where board-level policy fails to effectively mitigate supply chain risks. These findings suggest that financial institutions must pivot from volume-based compliance to risk-optimized integration to bridge these strategic and operational gaps. Full article

The increasing global demand for sparkling wines has encouraged the exploration of alternative grape varieties and emerging production regions. This study evaluated the potential of three indigenous Romanian grape varieties (Fetească regală, Tămâioasă românească, and Fetească albă) for sparkling wine production using the méthode champenoise, with grapes sourced from the ullu Mare region. The wines were characterized at two aging intervals (9 and 36 months on lees), with analyses performed on both disgorged and undisgorged samples to assess changes in physicochemical parameters, color attributes, volatile composition, and sensory properties. All varieties exhibited relatively high acidity (6.12–6.53 g/L), particularly Fetească regală (6.37–6.53 g/L), supporting their suitability for sparkling wine production. Extended lees aging enhanced the development of complex tertiary and quaternary aromas while preserving intrinsic floral and fruity attributes. Volatile analysis revealed aging-related increases in higher alcohols and medium-chain fatty acids, with 1-pentanol reaching 106.8 mg L⁻¹ and octanoic acid increasing from approximately 4.2 to 7.9 mg L⁻¹ after 36 months. Principal component analysis explained over 70% of the total variance, discriminating wines according to grape variety and maturation time. This study aimed to provide a detailed characterization of these sparkling wines, integrating physicochemical, chromatic, volatile, and sensorial analyses to evaluate their quality and enological potential. Full article

Background/Objectives: Rotavirus remains a major cause of severe acute gastroenteritis
in infants worldwide. The suboptimal efficacy of current vaccines underscores the need
for alternative microbiome-based interventions, including postbiotics. Extracellular
vesicles (EVs) from probiotic and commensal E. coli strains have been shown to mitigate
diarrhea and enhance immune responses in a suckling-rat model of rotavirus infection.
Here, we investigate the regulatory mechanisms activated by EVs in rotavirus-infected
enterocytes. Methods: Polarized Caco-2 monolayers were used as a model of mature
enterocytes. Cells were pre-incubated with EVs from the probiotic E. coli Nissle 1917 (EcN)
or the commensal EcoR12 strain before rotavirus infection. Intracellular Ca²⁺
concentration, ROS levels, and the expression of immune- and barrier-related genes and
proteins were assessed at multiple time points post-infection. Results: EVs from both
strains exerted broad protective effects against rotavirus-induced cellular dysregulation,
with several responses being strain-specific. EVs interfered with viral replication by
counteracting host cellular processes essential for rotavirus propagation. Specifically, EV
treatment significantly reduced rotavirus-induced intracellular Ca²⁺ mobilization, ROS
production, and COX-2 expression. In addition, both EV types reduced virus-induced
mucin secretion and preserved tight junction organization, thereby limiting viral access
to basolateral coreceptors. Additionally, EVs enhanced innate antiviral defenses via
distinct, strain-dependent pathways: EcN EVs amplified IL-8-mediated responses,
whereas EcoR12 EVs preserved the expression of interferon-related signaling genes.
Conclusions: EVs from EcN and EcoR12 act through multiple complementary
mechanisms to restrict rotavirus replication, spread, and immune evasion. These findings
support their potential as effective postbiotic candidates for preventing or treating
rotavirus infection. Full article

Background/Objectives: Malaria remains a major global health burden, increasingly complicated by resistance to artemisinin-based therapies. Because artemisinin activation depends on heme and porphyrin chemistry, we sought to exploit host red blood cell (RBC) heme metabolism as a therapeutic vulnerability. This study aims to develop and evaluate a host-directed “bait-and-kill” strategy that selectively sensitizes malaria-infected RBCs to artemisinin. Methods: We integrated quantitative proteomics, erythropoiesis transcriptomic analyses, flow cytometry, and in vitro malaria culture assays to characterize heme metabolism in mature RBCs and Plasmodium falciparum-infected RBCs (iRBCs). The heme precursor 5-aminolevulinic acid (ALA) was used to induce porphyrin accumulation, and dihydroartemisinin (DHA) was applied as the killing agent. Drug synergy, porphyrin accumulation, reactive oxygen species (ROS) induction, and parasite survival were assessed, including ring-stage survival assays using artemisinin-resistant clinical isolates. Results: Mature RBCs retain a truncated heme biosynthesis pathway capable of accumulating porphyrin intermediates, while uninfected RBCs are impermeable to ALA. In contrast, iRBCs exhibit increased membrane permeability, allowing selective ALA uptake and porphyrin accumulation. ALA alone did not induce cytotoxicity or ROS, whereas DHA induced ROS and parasite killing. The ALA + DHA combination resulted in synergistic parasite elimination, including complete clearance of artemisinin-resistant P. falciparum isolates from the Greater Mekong Subregion, with no recrudescence observed over three weeks of culture. Evidence supports a predominant role for host-derived heme metabolites in mediating this synergy. Conclusions: The bait-and-kill strategy selectively exploits host RBC heme metabolism to restore and enhance artemisinin efficacy while sparing uninfected cells. Using clinically safe compounds, this host-directed approach provides a promising, resistance-bypassing framework for malaria treatment and combination drug development. Full article

Artificial intelligence (AI) is now embedded across many aspects of healthcare, yet most implementations remain fragmented, task-specific, and layered onto legacy workflows. This paper does not review AI applications in healthcare per se; instead, it examines what an AI-first healthcare system would look like, one in which AI functions as a foundational organizing principle of care delivery rather than an adjunct technology. We synthesize evidence across ambulatory, inpatient, diagnostic, post-acute, and population health settings to assess where AI capabilities are sufficiently mature to support system-level integration and where critical gaps remain. Across domains, the literature demonstrates strong performance for narrowly defined tasks such as imaging interpretation, documentation support, predictive surveillance, and remote monitoring. However, evidence for longitudinal orchestration, cross-setting integration, and sustained impact on outcomes, costs, and equity remains limited. Key barriers include data fragmentation, workflow misalignment, algorithmic bias, insufficient governance, and lack of prospective, multi-site evaluations. We argue that advancing toward AI-first healthcare requires shifting evaluation from accuracy-centric metrics to system-level outcomes, emphasizing human-enabled AI, interoperability, continuous learning, and equity-aware design. Using hypertension management and patient journey exemplars, we illustrate how AI-first systems can enable proactive risk stratification, coordinated intervention, and continuous support across the care continuum. We further outline architectural and governance requirements, including cloud-enabled infrastructure, interoperability, operational machine learning practices, and accountability frameworks—necessary to operationalize AI-first care safely and at scale, subject to prospective validation, regulatory oversight, and post-deployment surveillance. This review contributes a system-level framework for understanding AI-first healthcare, identifies priority research and implementation gaps, and offers practical considerations for clinicians, health systems, researchers, and policymakers. By reframing AI as infrastructure rather than isolated tools, the AI-first approach provides a pathway toward more proactive, coordinated, and equitable healthcare delivery while preserving the central role of human judgment and trust. Full article

Pod corn (Zea mays var. tunicata) bears leafy glumes that enclose kernels, resembling a partial reversion to wild-forms, yet remains poorly characterized in situ in Mexico. We evaluated Mexican accessions at two contrasting locations to quantify morphological/phenological diversity and to assess functional traits via proximate kernel composition. Standard descriptors captured variation in plant architecture, tassel/ear traits (including glume length), and reproductive timing. Accessions showed strong plasticity and significant accession × environment effects on ear morphology and maturation. Grain yield ranged from 6.32 to 10.78 t ha⁻¹, with peak values comparable to commercial hybrids and above-typical yields reported for native Mexican races (2.7–6.6 t ha⁻¹). Proximate analysis showed that milling with the tunic increased moisture/ash (up to 3.07% vs. 1.80% in dehulled grain), tended to lower fat and protein, and yielded lower crude fiber than dehulled samples (0.78–0.96% vs. 1.59–1.77%); protein varied widely (1.05–6.64%). Thus, the tunic modulates elemental composition, informing processing choices (with vs. without tunic). Our results document a spectrum of morphotypes and highlight developmental diversity and field adaptability. The observed accession × environment responses provide a practical baseline for comparisons with native and improved varieties, and help guide product development strategies. Collectively, these data underscore the high productive potential of pod corn (up to 10.78 t ha⁻¹ under optimal management) and show that including the tunic substantially alters proximate composition, establishing a quantitative foundation for genetic improvement and food applications. Overall, pod corn’s distinctive ear morphology and context-dependent composition reinforce its value for conservation, developmental genetics, and low-input systems. Full article

Optimizing immunosuppressant dosing presents significant challenges in kidney transplantation due to narrow therapeutic ranges and considerable inter-patient pharmacokinetic differences. Emerging strategies for precision dosing, encompassing Bayesian population pharmacokinetic models, pharmacogenomic integration, and artificial intelligence algorithms, aim to enhance drug monitoring by moving beyond traditional trough-based approaches. This review critically assesses available evidence for predictive dosing models targeting immunosuppressants, including calcineurin inhibitors, antimetabolites, and mTOR inhibitors in kidney transplant patients. Available observational and simulation studies demonstrate substantial methodological diversity, with Bayesian PopPK-guided strategies showing 15–35% better target exposure achievement compared to trough-based monitoring. The absence of pooled estimates precludes a precise summary effect size, and evidence from randomized controlled trials remains limited. Machine learning models, particularly for tacrolimus, frequently reduced prediction error relative to traditional regression approaches, but substantial heterogeneity in study design, outcome definitions, and external validation limits quantitative synthesis. Hybrid Bayesian–AI frameworks and explainable AI tools show conceptual promise but are largely supported by proof-of-concept studies rather than reproducible clinical implementations. Overall, Bayesian pharmacokinetic modelling represents the most mature and clinically interpretable approach for precision dosing in transplantation, whereas AI-driven and hybrid systems remain investigational. Key gaps include the need for standardized reporting, rigorous risk-of-bias assessment, prospective validation, and clearer regulatory and implementation pathways to support safe and equitable clinical adoption. Full article