Enterprise systems and intelligent systems have evolved along separate paths, resulting in a fragmentation between mathematically rigorous theories of intelligence and practically proven methodologies for system engineering. This paper introduces the Constrained Zoned‑Object Architecture (CZOA), a unified formalism that integrates the theoretical depth of Constrained Object Hierarchies (COH) with the engineering pragmatism of the Zoned Role‑Based (ZRB) framework. We demonstrate that enterprise systems are a natural species of intelligent systems and that the COH 9‑tuple can be instantiated within ZRB's zone‑role structure to create systems that are simultaneously secure, intelligent, maintainable, and organizationally coherent. The integrated framework preserves minimality while enabling new capabilities: neural‑enhanced permission mining, semantic embeddings for cross‑zone understanding, adaptive access control, and continuous daemon‑based enforcement. Category‑theoretic foundations support compositional reasoning, and the UniLang logic language provides formal constraint specification. The CZOI toolkit implements CZOA in Python, offering a modular foundation for building secure and intelligent integrated organizational systems. Through five condensed case studies spanning healthcare, finance, traffic management, higher education, and supply chain, we provide quantitative evidence of the framework's effectiveness, including a 23% reduction in wait times during surge scenarios, 94% anomaly detection accuracy, and sub‑millisecond permission decisions. The paper concludes with a discussion of limitations, threats to validity, and future research directions.

Bietti crystalline dystrophy (BCD) is a hereditary retinal disease caused by loss-of-function mutations in the CYP4V2 gene. Gene replacement therapy using rAAV-hCYP4V2 represents a promising therapeutic strategy, requiring robust bioassays for product quality control. This study developed and validated a sensitive LC-MS/MS method for quantifying CYP4V2 enzyme activity. Lysates from HeLa-AAVR cells transduced with rAAV-hCYP4V2 (MOI=3×105) were used, with lauric acid as substrate supplemented with cytochrome P450 reductase, cytochrome b5, and NADPH. The ω-hydroxylated product (12-hydroxy lauric acid) was quantified using tolbutamide as internal standard. Method validation followed ICH guidelines. Results demonstrated excellent specificity with negligible background in negative controls. Linearity was achieved over 0.5-100 ng/mL (R2>0.99), with average recovery of 100.6%. Intra-batch and inter-batch precision RSDs were <47.8% and <28.4%, respectively. Product stability was maintained ≥4 weeks at -80℃. The method was successfully applied to three AAV serotypes (AAV2, AAV8, AAV2/8), with all RSDs <23.9%. This validated LC-MS/MS bioassay provides a crucial quality control tool for potency assessment, process development, batch release, and stability studies of rAAV-hCYP4V2 gene therapy products.

Polymorphic adversarial agents that dynamically mutate their behavior present significant challenges to conventional intrusion detection systems, which often rely on static feature representations or fixed signature based models. Existing deep learning approaches, including CNNs, transformers, and graph neural networks (GNNs), demonstrate strong detection capabilities but exhibit limitations in handling continuously evolv ing attack patterns. In this paper, we propose an integrated entropy-driven agentic cyber defense model that combines graph neural network-based anomaly detection, generative adversarial network (GAN) poly morphic attack simulation, reinforcement learning-based mitigation, and game-theoretic attacker–defender modeling. It models the stochastic mutation of adversarial agents using entropy gradients, captures relational network structures through GNNs, simulates realistic polymorphic attacks using GANs, and dynamically adapts mitigation strategies via reinforcement learning. Experimental evaluation on the CICIoT2023 dataset demonstrates that the proposed model achieves superior performance over existing state-of-the-art methods, with an accuracy of 98.3%, F1-score of 0.98, and significant improvements in robustness against polymor phic attacks compared to CNN-based (93.2% accuracy, 0.92 F1), transformer-based (95.4%, 0.94 F1), and conventional GNN-based (96.1%, 0.95 F1) intrusion detection systems. Additional ablation studies confirm the contribution of entropy modeling, GNN embeddings, and RL-based mitigation to overall system effec tiveness. Future work will explore federated and distributed cyber defense architectures, integration with edge computing for IoT environments, and adaptive policy learning under large-scale network conditions, enabling real-time resilience against highly sophisticated adaptive cyber threats.

Soft tissue reconstruction often requires biomaterials that provide temporary mechanical support while allowing vascular integration and tissue remodeling. In reconstructive breast surgery, these demands converge within a uniquely challenging environment characterized by large surface areas, variable perfusion, frequent exposure to radiation, and reliance on prosthetic implants. As a result, breast reconstruction has emerged as a clinically relevant model for evaluating the performance and limitations of soft tissue scaffolds. Acellular dermal matrices (ADM) were initially adopted to provide biologically derived reinforcement based on the premise of host integration and neovascularization. While ADM reshaped implant-based reconstruction, accumulating clinical experience has revealed important constraints, including variability in mechanical properties, inconsistent vascularization, susceptibility to fibrosis, and limited performance in compromised tissue beds. These limitations have driven increasing interest in synthetic polymer scaffolds engineered for predictable mechanics, controlled degradation, and scalable manufacturing. This narrative review examines the evolution from ADM to synthetic and hybrid scaffold systems in breast reconstruction. We discuss how scaffold architecture, thickness, porosity, and degradation kinetics influence angiogenesis, immune response, and mechanical load transfer during healing. Hybrid strategies that integrate selective bioactivity within synthetic frameworks are also considered, highlighting both their translational promise and practical challenges. These concepts are particularly relevant for implant-based breast reconstruction, where scaffold performance directly influences complication rates, implant stability, and long-term reconstructive outcomes.

The evolution of pulsed discharge behavior inside a preformed air bubble in water from the first to subsequent pulses was experimentally investigated using a synchronized needle–bubble system. A positive nanosecond high-voltage pulsed power supply, together with a pulse valve and ICCD imaging, was employed to generate reproducible preformed bubbles and to record the corresponding discharge development with good temporal synchronization. The results show that, although the preformed bubbles exhibit good repeatability in size and morphology under identical conditions, the first-pulse discharge inside the bubble remains highly stochastic. The first discharge is predominantly corona-like and is not significantly affected by bubble size once the electrode is covered by the bubble. By varying the pulse width, the discharge inside the bubble was observed to evolve progressively from corona-like emission to streamer discharge, accompanied by increasing instability of the bubble interface. At sufficiently large pulse width and pulse number, bubble wrinkling and even rupture were induced. The effect of solution conductivity was also examined. Increasing conductivity significantly enhanced discharge intensity, enlarged the luminous region, and promoted streamer propagation along the inner bubble surface. At sufficiently high conductivity, the first pulse already produced strong discharge and rapid bubble rupture. In addition, the current amplitude and the energy dissipated per pulse increased with conductivity and pulse number. These results demonstrate that the discharge evolution inside a preformed bubble is jointly governed by pulse history, pulse width, and solution conductivity, and that residual effects from previous pulses play an important role in the transition from the first pulse to subsequent discharges.

Time–frequency decompositions, statistical descriptors and nonlinear dynamical methods can be used to assess electroencephalographic (EEG) signals. These approaches treat signals as time series evolving in Euclidean state spaces, potentially overlooking global geometric constraints. We explore an alternative representation in which EEG dynamics are embedded in a Möbius-like manifold, a non-orientable geometric structure capable of encoding cyclic evolution and phase-dependent state inversion within a compact topological space. We combine normalized signal amplitude with instantaneous phase derived from the Hilbert transform to generate a three-dimensional trajectory representing the temporal evolution of neural electrical activity. Using EEG recordings from a healthy young adult, we reconstructed trajectories and computed path geometric descriptors, including torsion and cumulative winding number. These two quantities characterize local twisting and global rotations, providing information about the trajectories of oscillatory neural activity. We derive testable hypotheses from our proof-of-concept study, e.g., transitions between oscillatory regimes should appear as changes in torsion or winding structure. We conclude that embedding neural signals in non-orientable manifolds could provide a methodological framework for assessing physiological and pathological brain activity.

Language agents that act through tools, files, browsers, APIs, and persistent sessions are shaped by more than the base model or a single prompt. Their reliability depends on a harness layer that determines which instructions remain authoritative, what actions are available, how state is carried forward, and how failures are handled over time. This position paper argues that recent practice has made this layer visible enough to warrant explicit treatment in NLP. We propose and operationalize a working decomposition of the harness layer as control, agency, and runtime (CAR); situate harness engineering in the arc from software engineering through prompt and context engineering; and provide a lightweight audit of 40 harness-relevant works in our selected evidence base, suggesting a visibility gap between academic papers and public engineering notes. We further argue that many reported agent gains may be partly harness-sensitive rather than purely model-driven, and propose HarnessCard as a lightweight reporting artifact, including a filled example. Grounded in papers, benchmarks, protocols, and engineering notes through 20th Mar, 2026, we argue that progress in language agents should report not only the model, but also the harness layer that turns capability into governed action.

Highly pathogenic avian influenza (HPAI) H5N1 clade 2.3.4.4b continues to circulate extensively among wild birds, poultry, and mammals, presenting ongoing risks at the intersection of human, animal, and environmental health. Antiviral approaches tailored for poultry farming or farm settings are still largely under investigation. The influenza A polymerase acidic (PA) endonuclease, which plays a key role in cap-snatching during viral transcription, is a conserved antiviral target across different host species. This research introduces a computational workflow to detect PA endonuclease inhibitors suitable for veterinary and environmental use. Homology modelling of recent H5N1 clade 2.3.4.4b PA sequences (2021–2025) was performed based on the crystallographic structure 6FS8. A curated ligand library, including reference inhibitors, hydrophilic metal-binding scaffolds, and repurposed antivirals, was screened via cross-host molecular docking against poultry- and mammalian-specific PA models. Docking results highlighted baloxavir and entecavir as top candidates due to their strong and consistent binding profiles across multiple targets. Entecavir showed particularly promising binding affinity in the poultry PA model (−100.6), similar to the reference inhibitor baloxavir (−97.5 to −97.7). A 170 ns molecular dynamics simulation of the poultry PA–entecavir complex indicated stable structural behavior, with RMSD below 1.1 Å, and MM/PBSA calculations gave a binding free energy of ΔG = −85.1 ± 0.8 kJ/mol. Physicochemical analysis revealed entecavir's high polarity and predicted water solubility, suggesting its suitability for water-based or environmental delivery within poultry facilities. Overall, this study introduces a One Health–focused computational framework that integrates cross-host structural modeling, docking, molecular dynamics, and agrochemical suitability filtering to identify promising antiviral candidates for poultry outbreak control. The results point to entecavir as a promising candidate for further evaluation in veterinary antiviral studies.

This study analyzes the vibrations to which the WBV of an agricultural tractor operator is exposed during the performance of three different tillage’s: Standard Tillage (ST), System Deep (CTD) and System Shallow (CTS). Measurements were carried out according to ISO 2631-1 and ISO 2631-4 in three coordinate axes (x, y, z). The analysis and statistical processing carried out indicates that none of the mean vibration values exceeds the prescribed limit value of 1.15 m/s² according to Directive 2002/44/EC, but exceeds the daily warning value of 0.5 m/s². From the above, it can be assumed that the operator of the investigated tractor may be exposed to the occurrence of occupational diseases by long-term exposure to such vibrations. The highest vibrations in the x and y axes were recorded when working with the ripper, while in the z-axis they were the highest in the loosener. Although the measured values are within safe (permissible) limits, further research on this topic and optimization of operation can further reduce the burden on the operator.

Artificial intelligence (AI) is increasingly integrated into dental diagnostics, treatment planning, documentation, and research. While ethical principles such as transparency and accountability are widely discussed, there is limited synthesis of how informed consent should be conceptualized and operationalized within the evolving dentist–patient–AI relationship. This scoping review aimed to map existing evidence on informed consent in AI-augmented dentistry and dental research, identify conceptual and practical gaps, and propose a structured framework to support ethically robust implementation; Methods: PRISMA-ScR guidelines was followed with review question formulated using the Population–Concept–Context (PCC) framework. A systematic search was conducted in PubMed, Web of Science, and ClinicalKey, complemented by grey literature screening; Results: From 2624 identified records, 30 studies were included after screening. The literature consistently emphasized disclosure of AI involvement, clarification of clinician accountability, communication of algorithmic limitations and bias, and separation between clinical and research consent. Based on thematic synthesis, we propose the ACCOUNT-AI framework, comprising structured domains addressing AI role clarification, clinician accountability, contextual differentiation, operational risks, secondary data governance, adaptive consent design, and transparency across the AI lifecycle. The framework integrates clinical use, research applications, and regulated data reuse as components of a unified accountability model; Conclusions: Informed consent in AI-augmented dentistry requires adaptation from traditional bilateral models to a triadic dentist–patient–AI framework grounded in human professional accountability. Standardized, context-sensitive consent structures are needed to ensure transparency, protect patient autonomy, and support ethically responsible AI integration in both clinical care and research.