Search Results (8,236)

The aim of this research is to present a risk-based AI assurance framework that produces
quantifiable metrics for auditors and stakeholders to make deployment decisions with
evidence-driven assurance of traceability, explainability, accountability, and reproducibility. Our proposed framework incorporates risk severity core with additional modifiers to
accommodate the context, governance obligations, technical and environmental exposure,
and residual risk relevant to the AI model. This multi-tiered technique enables stakeholders
and governance teams to operationalize the safe deployment assurance. The final Assurance
Adequacy Score (AAS) comprises a Governance Readiness Score (GRS) along with
two additional indices to quantify the traceability and explainability of the AI model. The
Traceability Adequacy Index (TAI) is calculated by evaluating the attributes such as the
dataset and model versioning, pipeline logging, model audit completeness, and reproducibility. And an Explainability Adequacy Index (EAI) is calculated by evaluating the
attributes such as the fidelity for local and global explanations, stability, faithfulness of the
explanation provided, robustness, coverage, and human comprehension. This architecture
enables integration of risk assessment and enables continued AI assurance by deploying
a bottleneck principle where the readiness of the AI model is confined by the weaker of
the indices. Finally, a tiered gate mechanism is applied on the Assurance Adequacy Score
to enforce minimum assurance floors for high-risk AI systems. The evaluation conducted
on multi-domain AI models demonstrates the Risk-Based AI Assurance Framework’s
(RBAAF) ability to yield stable and consistent readiness decisions with sensitivity analysis
and re-scoring. The use cases demonstrate that even comparable risk levels can lead
to significantly different deployment outcomes depending on assurance maturity, and
design-specific improvements in traceable or explainable domains have the ability to shift
gate outcomes. Combining governance regulations with a standardized and quantifiable
traceability and explainability score enables the stakeholders to evaluate the AI system for
an accountable and regulation-compliant deployment. Full article

Background: Pantoprazole is a widely used proton pump inhibitor that is highly unstable under acidic conditions. This limits the performance of conventional formulations and typically requires enteric-coated dosage forms or alternative modified-release approaches. This study reports the development of polymeric matrix mini-tablets designed to protect pantoprazole during gastric exposure and to enable pH-dependent release under intestinal conditions. The formulations combine Eudragit^® S 100, a pH-dependent polymer, with HPMC, a hydrophilic matrix former that modulates drug release through hydration and swelling. Methods: Matrix mini-tablets were prepared by blending pantoprazole with selected excipients at optimised proportions and compressing the blends by direct compression using an eccentric tablet press. Powder blends and mini-tablets were characterised according to pharmacopoeial specifications. Analytical techniques—including High-Performance Liquid Chromatography (HPLC), Differential Scanning Calorimetry (DSC), Fourier-Transform Infrared Absorption Spectroscopy (FT-IR), Powder X-Ray Diffraction (PXRD), and Scanning Electron Microscopy (SEM)—were employed to evaluate drug content uniformity, thermal behaviour, and potential drug–excipient interactions. In vitro dissolution studies were performed under sequential pH conditions, and the release kinetics were analysed using mathematical models. Results: Dissolution testing identified formulations F2 and F6 as providing the most suitable gastro-resistant performance in the acidic stage, together with sustained release up to 24 h. Kinetic modelling supported formulation-dependent release mechanisms, and multivariate analysis (PCA) highlighted relationships between physico-mechanical attributes and drug-release behaviour. Conclusions: The proposed matrix system shows potential as a robust, coating-free platform for the modified delivery of acid-labile drugs using direct compression, simplifying manufacturing. These findings support the rational design of oral modified-release formulations based on polymeric matrices. Full article

Objectives: This study systematically evaluated and quantitatively synthesized preclinical evidence on the effects of magnesium (Mg) incorporation into or coating of titanium dental implants on osseointegration and peri-implant bone formation. Methods: Electronic searches of PubMed, Scopus, and Web of Science were performed up to May 2025 to identify animal studies evaluating Mg-modified titanium implants. Eligible studies compared Mg-incorporated or Mg-coated implants with non-modified titanium controls and reported quantitative histomorphometric outcomes. Primary outcomes included the values of bone-to-implant contact (BIC) and bone area (BA) around implants. Study quality was assessed using the ARRIVE 2.0 guidelines. Meta-analyses were performed using weighted mean differences with 95% confidence intervals under fixed- or random-effects models based on heterogeneity. Results: Eleven preclinical animal studies conducted in rabbit and rat models were included. Mg was incorporated using various surface-modification techniques, including ion implantation, Mg-substituted hydroxyapatite coatings, mesoporous titania layers, and nanotubular structures. Overall, the studies’ quality was high, with most studies rated as excellent and with a low-to-moderate risk of bias. Furthermore, the meta-analysis revealed a significant increase in BIC for Mg-modified implants compared with uncoated implants (Z = 4.38, p < 0.001), implying improved osseointegration. Meanwhile, pooled BA values showed no significant differences between the groups (Z = 0.93, p = 0.35). Conclusions: Mg coating onto or incorporation into titanium implant surfaces can improve BIC in preclinical models, indicating improved osseointegration in the early stages. Full article

Sulfonamide antibiotics (SAs) are ubiquitous and persistent organic contaminants in aquatic and soil ecosystems due to their extensive application and high structural stability, causing rising environmental hazards. Conventional treatment approaches, generally based on physical adsorption or biological processes, remain limited in achieving efficient and stable removal as well as deep molecular modification of SAs. In recent years, modified biochar has developed as a flexible environmental functional material incorporating adsorption and reaction regulation capabilities, owing to its customizable pore structure, surface chemistry, and electronic characteristics. This study comprehensively highlights current achievements in the adsorption and degradation of sulfonamide antibiotics by modified biochar, with specific emphasis on modification techniques, structural modulation, structure–performance connections, and interfacial reaction processes. Through physical activation, heteroatom doping, defect engineering, and metal integration, biochar has developed from a traditional adsorbent into a carbon-based interfacial reactor capable of pollutant adsorption, molecular activation, and directed transformation. Surface-confined reaction interfaces, where π–π interactions, hydrogen bonding, electrostatic interactions, and metal coordination cooperatively control adsorption and transformation processes, are primarily responsible for the elimination of SAs. Moreover, the dual functions of modified biochar in driving both radical and non-radical pathways are explored, showing the vital importance of interfacial electronic structure modulation and electron-transfer mechanisms in influencing reaction efficiency and selectivity. The impact of sulfonamide molecular configurations, ambient circumstances, and concomitant chemicals on removal performance are also explored. Unlike previous reviews that mainly summarize adsorption efficiency or oxidant activation systems separately, this work integrates structural modulation, interfacial electronic regulation, and bond-selective transformation mechanisms into a unified structure–chemistry–reactivity framework. By correlating sulfonamide molecular configuration with biochar electronic structure, this review provides a mechanistic roadmap for the rational design of next-generation catalytic biochar systems. Finally, key challenges related to structural controllability, long-term stability, and engineering scalability are identified, and future research directions are proposed to support the rational design of high-performance biochar materials and the practical control of sulfonamide antibiotic pollution. Full article

This article evaluates the use of probing classifiers to modify the internal hidden state of a chess-playing transformer, which has been trained on sequences of chess moves and can generate new moves with prompted. Probing classifiers are a technique for understanding and modifying the operation of neural networks in which a smaller classifier is trained to use the model’s internal representation to learn a probing task. The aim of this research is to discover whether the learned model possesses an editable internal representation of the chess game, despite being trained without explicit information about the rules of chess. We contrast the performance of standard linear probes against Sparse Autoencoders (SAEs), a latent space interpretability technique designed to decompose polysemantic concepts into atomic features via an overcomplete basis. Our experiments demonstrate that linear probes trained directly on the residual stream significantly outperform probes based on SAE latents. When quantifying the success of interventions via the probability of legal moves, linear probe edits achieved an 88% success rate, whereas SAE-based edits yielded only 41%. These findings suggest that while SAEs are valuable for specific interpretability tasks, they do not enhance the controllability of hidden states compared to raw vectors. Finally, we show that the residual stream respects the Markovian property of chess, validating the feasibility of applying consistent edits across different time steps for the same board state. Full article

The influence of YB₄ particle addition on the microstructure and the associated thermal and mechanical properties of AISI 420 stainless steel composites fabricated using the vacuum induction melting technique was investigated. Microstructural analysis using scanning electron microscopy (SEM) confirmed the presence of YB₄ particles within the BCC-structured martensitic matrix and also along the grain boundaries across all weight fractions. In addition, YB₄ addition resulted in a pronounced refinement of the martensitic matrix, as evidenced by a progressive reduction in the size of the packets, i.e., a group of martensitic laths/plates sharing the same habit plane variants with the parent austenite grain. The presence of YB₄ particles induced internal stresses and microstrains, leading to peak shifting and broadening of the X-ray diffraction (XRD) peaks corresponding to that of the martensitic matrix phase. The coefficient of thermal expansion (CTE) decreased significantly from 13.4 × 10⁻⁶ K⁻¹ for monolithic AISI 420 to 8.06 × 10⁻⁶ K⁻¹ for the AISI 420/4 wt.% YB₄ composite and was attributed to the excellent dimensional stability of YB₄ particles. The maximum hardness (913.12 HV) and tensile strength (930 MPa) were achieved for the AISI 420/4 wt.% YB₄ composite. Fractographic analysis using SEM indicated a transition from ductile to brittle fracture with increasing YB₄ content, suggesting a reduction in strain-hardening capacity. The contributions of various strengthening mechanisms were quantified using the summation of strengthening and modified Clyne models, revealing that strengthening due to load bearing is dominant across all composites. Insights gained from these results are important to strategize the design of boride-based metal-matrix composites with enhanced strength–ductility synergy for structural applications. Full article

Ciprofloxacin (CIP) is an antibiotic that is widely used in humans and animals. However, the compound has been detected in animal-derived products and the environment due to its extensive use, causing serious concern for public health and environmental safety. The issue raises the urgent need to develop innovative techniques to monitor CIP. Therefore, this study aims to develop a simple and sensitive CIP sensor called the boron-doped diamond nanoparticle-modified screen-printed electrode (BDD NPs/SPE) and the nickel oxide nanoparticle-modified BDD NPs/SPE (NiO NPs/BDD NPs/SPE). NiO NPs were synthesized via green synthesis using Spatholobus littoralis Hassk. root extract as the reducing agent. The formation and characteristics of NiO NPs were then confirmed through a UV-Vis spectrophotometer, XRD, PSA, FT-IR, and XPS. The successful modification of SPE was confirmed through SEM-EDX, followed by measurements using square-wave voltammetry. The results showed that the modified SPE could detect CIP over a concentration range of 0.1–100 µM and produced a low detection limit of 0.109 µM for BDD NPs/SPE and 0.054 µM for NiO NPs/BDD NPs/SPE. The proposed method was successfully applied to the determination of CIP in commercial tablets, milk, and human urine, with a satisfactory % recovery from 95 to 100%. The current study successfully developed a simple yet highly sensitive sensor that enabled robust, reliable, and efficient detection of CIP, showing its strong potential for practical applications. Full article

Titanium alloys are widely used in biomedical applications, especially in dental implants. In this work, individual TiO_xC_y thin films and a novel multilayer coating approach were investigated to prevent early implant failure through surface properties optimization. The research focuses on designing an innovative TiO_xC_y organometallic multilayer coating, varying from mineral (low C) to organic (high C), on Ti6Al4V substrates. These coatings were prepared using the PECVD technique, varying parameters as the reactive gas flow to modify the chemical composition, hydrophilicity, and layer thickness. Comprehensive characterization of the surface was conducted using XPS, and by contact angle to evaluate wettability. To further understand the chemical composition within each layer, XPS depth profiling analyses were performed. The results revealed that the newly designed multilayer coating with a decreasing reactive gas flow clearly exhibited a gradient in its composition. Near the upper substrate surface, the layers display a mineral-like, low-carbon structure, transitioning to an organic-like, high-carbon composition at the outermost surface. Full article

This study presents the synthesis and ball-milling modification of titanite (CaTiSiO₅) powders to enhance the thermal stability and performance of polypropylene (PP) separators for lithium-ion batteries (LIBs). CaTiSiO₅ was synthesized using a ceramic route, and the experimental design varied the milling cycles and sphere sizes. Characterization techniques, including scanning electron microscopy, X-ray diffraction, Fourier-transform infra-red spectroscopy, surface area analysis, thermal analysis, and electrochemical tests, confirmed the production of high-purity monoclinic CaTiSiO₅. Ball milling effectively reduced the particle and crystallite sizes while increasing the specific surface area, total pore volume, double-layer capacitance, and ionic conductivity, while also reducing the cell resistance. Coating PP separators with the modified CaTiSiO₅ significantly improved their thermal stability and enhanced their electrochemical properties, including the electron transfer rate and Coulombic efficiency. These findings demonstrate the potential of ball-milled CaTiSiO₅ as a valuable material for developing safer and more efficient LIBs. Full article

The state budget system in Vietnam functions within a cohesive structure that allocates financial resources between central and local governments; nevertheless, substantial disparities in socioeconomic conditions among provinces have resulted in increasing discrepancies in local budget revenue. This study, therefore, examines the impacts of fiscal decentralization policy, land utilization, urbanization, provincial competitiveness index, and human capital on local government revenue. The analysis utilizes quantitative panel-data techniques on a dataset encompassing all 63 Vietnamese provinces and municipalities from 2017 to 2022, totaling 378 observations. Econometric estimation employs pooled ordinary least squares, fixed-effects, random-effects, and viable generalized least squares models, along with diagnostic and robustness checks to mitigate unobserved heterogeneity and error dependence. The findings demonstrate statistically significant correlations between local budget revenue and five studied determinants. However, fiscal decentralization policy exerts the most significant influence on the revenue of the local government budget. The results suggest that enhancing municipal fiscal performance needs more than merely modifying revenue-sharing ratios, with significant ramifications. Full article

As deep learning models become increasingly integrated into critical decision-making systems, the need for explainable Artificial Intelligence (xAI) has grown paramount to ensure transparency, accountability, and trust. Post hoc explainability methods, which analyse trained models to interpret their predictions without modifying the underlying architecture, have become increasingly important, especially in fields such as healthcare and finance. Modern xAI techniques often produce feature importance rankings that fail to capture the true causal influence of features, particularly in transformer-based models. Recent quantitative metrics, such as Symmetric Relevance Gain (SRG), which measures the area between the feature corruption performance curves of the Most Important Feature (MIF) and the Least Important Feature (LIF), provide a more rigorous basis for evaluating explanation fidelity. In this study, we first show that existing xAI methods exhibit consistently poor performance under the SRG criterion when explaining transformer-based text classifiers. To address these limitations, we introduceEvoDropX, a novel framework that formulates explanation as an optimisation problem. EvoDropX leverages Grammatical Evolution (GE) to evolve sequences of feature corruption with the explicit objective of maximising SRG, thereby identifying features that most strongly influence model predictions. EvoDropX provides interventional, input–output (behavioural) explanations and does not attempt to infer or interpret internal model mechanisms. Through comprehensive experiments across multiple datasets (IMDb movie reviews (IMDB), Stanford Sentiment Treebank (SST-2), Amazon Polarity (AP)), multiple transformer models (Bidirectional Encoder Representations from Transformers (BERT), RoBERTa, DistilBERT), and multiple metrics (SRG, MIF, LIF, Counterfactual Conciseness (CFC)), we demonstrate that EvoDropX significantly outperforms all state-of-the-art (SOTA) xAI baselines including Attention-Aware Layer- Wise Relevance Propagation for Transformers (AttnLRP), SHapley Additive exPlanations (SHAP), and Local Interpretable Model-agnostic Explanations (LIME), when evaluated using intervention-based faithfulness criteria. Notably, EvoDropX achieves $74.77\%$ improvement in SRG than the best-performing baseline on the IMDB dataset with the BERT model, with consistent improvements observed across all dataset-model pairs. Finally, qualitative and linguistic analyses reveal that EvoDropX captures both sentiment-bearing terms and their structural relationships within sentences, yielding explanations that are both faithful and interpretable. Full article

We present a high-resolution 3-D shear-wave velocity model of the Indonesian lithosphere and upper mantle, constructed through a weighted joint inversion of complementary surface wave datasets. Our model integrates teleseismic Rayleigh waves from 387 earthquakes recorded at 31 stations, analyzed using a modified two-plane-wave tomography method, with two years of ambient noise data from 30 stations processed via image transformation techniques. Our results provide new structural constraints on the four principal subduction systems in Indonesia. Along the Sunda–Java Trench, the slab exhibits a systematic along-strike transition from a continuous and well-defined geometry in the west to increasingly disrupted and thickened structures toward the east. This evolution correlates with the subduction of progressively older lithosphere. Beneath the Banda Arc, we image a continuous slab whose dramatic 180° curvature and deep coalescence of distinct segments provide direct evidence for a single-slab rollback and folding origin. In the Molucca Sea region, tomography reveals a shallow low-velocity zone and resolves the complex geometry of an active double-sided subduction system associated with arc–arc collision. Collectively, these findings provide unprecedented constraints on slab segmentation and deformation, highlighting the dominant control of lithospheric age and complex plate interactions on the geodynamic evolution of this exceptional convergent boundary. Full article

In this work, a novel nanostructured Mn₃O₄-based electrochemical sensor was developed for the determination of heavy metals in aqueous media. The Mn₃O₄ nanostructure was solvothermally synthesized in the sole presence of propylene glycol (PG). Under the specific synthetic conditions, PG provided surface coating and stabilization by decomposition products and/or residual PG molecules that have been adsorbed on Mn₃O₄ NPs surfaces, creating a thin organic layer. This imparts a negative surface charge (zeta potential), enhancing colloidal stability in dispersions and electrochemical performance. The physicochemical properties of the resulting NPs were characterized via X-ray diffraction (XRD), Fourier transform infrared (FT-IR), Thermogravimetric Analysis (TGA), and Dynamic light scattering (DLS) and ζ-potential measurements, as well as SEM imaging of the modified electrode surface, confirming its successful formation and favorable structural properties. The LODs of Cd²⁺, Pb²⁺, Zn²⁺, and Cu²⁺ for their simultaneous determination are 2.9 μg·L⁻¹, 5.2 μg·L⁻¹, 7.1 μg·L⁻¹, and 2.5 μg·L⁻¹, respectively, with relative standard deviations of about 5.24%, 4.43%, 7.74%, and 4.53%, respectively. As a result of this study, a simple, sensitive, and reproducible electrochemical sensor based on a carbon paste electrode (CPE) modified with novel synthesized manganese nanoparticles and employing voltammetric techniques was applied in water and wastewater. Full article

Background/Objectives: The use of the drug delivery system is notable for the systemic improvement of low orally bioavailable compounds, such as the bioactive phenolic acid, syringic acid. Innovative techniques are employed to enhance the performance of certain drug delivery systems. In connection with our previously reported journal with the use of MIL-100(Fe) as a drug carrier for syringic acid, this study utilized a mixed-linker synthesis of syringic acid and trimesic acid and characterized the properties in comparison with the unmodified MIL-100(Fe) through a solid solution approach. Methods: Modified MIL-100(Fe) was synthesized by substituting different molar concentrations of syringic acid for trimesic acid through de novo synthesis. Simple impregnation of syringic acid was carried out at 12, 24, 36, and 48 h and at 1:1 and 1:2 molar ratios of MIL-100(Fe) to syringic acid. Characterization was performed via PXRD, FTIR, BET, SEM, and DLS. In vivo studies included acute oral toxicity testing (OECD 425) and bioavailability assessment in Sprague Dawley rats. Results: The optimized amount of syringic acid to be substituted for trimesic acid is 0.10 mmol, as confirmed by the value of the PXRD. Optimized drug loading of 66.85 ± 0.004% was achieved using a 1:2 ratio of syringic acid to MIL-100(Fe)-10% over 36 h. Structural modifications were confirmed via FTIR, specifically through shifts at 1239.2 cm⁻¹, while TGA demonstrated thermal stability up to approximately 350 °C. Morphological analysis by SEM showed octahedral particles (210.70 ± 1.23 nm), and a decrease in BET surface area post-loading verified successful encapsulation. While in vitro release was media-dependent, toxicity studies at 2000 mg/kg showed no adverse effects; notably, SGOT and SGPT levels decreased, though BUN and creatinine levels rose. Compared to pure oral syringic acid, the SYA@MIL-100(Fe)-10% formulation demonstrated a 5.09-fold increase in relative bioavailability. Furthermore, it outperformed intraperitoneal administration of the drug by 1.65-fold. Conclusions: Modification of MIL-100(Fe) by incorporating syringic acid into the framework as a substituted organic linker indicates that SYA@MIL-100(Fe)-10% is a safe and effective delivery system for syringic acid, enhancing oral bioavailability. To the best of our knowledge, this is the first study to investigate the mixed-linker synthesis of MIL-100(Fe) by utilizing syringic acid as a structural co-ligand, rather than solely as an encapsulated guest. While MIL-100(Fe) has been extensively employed as a carrier for various therapeutics, this research uniquely integrates the active agent into the framework lattice itself to modulate porosity and loading capacity, subsequently evaluating its systemic performance in an in vivo model. Full article

Background/Objectives: The growing demand for personalised, patient-centric drug delivery systems has driven innovation in pharmaceutical manufacturing, particularly in multi-unit particulate systems (MUPS). Methods: In this study, inert cores with tailor-made geometry for multi-particulate formulations were fabricated with high-resolution stereolithography (SLA) 3D printing. By a printable photopolymer resin, dimensionally accurate and mechanically robust starter cores were produced. The additively manufactured inert subunits were drug-layered with ibuprofen sodium using a fluidised bed process. Then, a controlled-release film coating of Eudragit RS 30D was applied with varying coating thicknesses. The initial 3D-printed subunits, together with the drug-layered and finally film-coated microparticles, were characterised by image analysis, Raman microspectroscopic measurements, and official methods of the European Pharmacopoeia. Results: The combined approach of 3D printing and traditional pharmaceutical processing proved highly effective. The 3D-printed cores demonstrated both flexibility in design and consistency in performance. Conclusions: These findings highlight the feasibility of using 3D printing to produce patient-specific, functional cores in multi-particulate systems that can be easily modified according to the patient’s needs. The fabricated minitablets can be used as alternatives to widely used inert cores. Integrating additive manufacturing with conventional coating techniques offers promising new avenues for developing next-generation, personalised drug delivery solutions. Full article